Building Real Time Targeting Capabilities - Ryan Zotti, Subbu Thiruppathy - Capital One

•Download as PPTX, PDF•

5 likes•776 views

The document discusses building real-time targeting capabilities at Capital One. It introduces two speakers, Ryan Zotti and Subbu Thiruppathy, and describes challenges around striving for speed in everything. It then covers how to achieve fast model data, training, deployment, and scoring through techniques like using the most up-to-date data, distributed computing in the cloud, automatic model refitting, and response times under 100 milliseconds.

Building Real Time Targeting Capabilities
Capital One | Fast Marketing
July 20, 2016 | H20 Open Tour | NYC

Ryan Zotti
Senior Data Engineer
Subbu Thiruppathy
Senior Software Engineer
EXPERTISE
FUN FACT
Big Data, Python, R, Java,
Machine Learning, AWS
FUN FACT
Big Data, Java, AKKA
Play, AWS
Built a self driving
remote controlled car
Recipient of Capital One’s
most prestigious honor
EXPERTISE

Building Real Time Targeting Capabilities - Ryan Zotti, Subbu Thiruppathy - Capital One

http://www.tibco.com/blog/2015/05/26/upcoming-webinar-integration-as-the-foundation-of-fast-data-may-28-2/

Source: http://thumbs.dreamstime.com/x/65-miles-per-hour-7772157.jpg
https://morganalyx.wordpress.com/2013/02/22/assertive-driving/

http://www.dailymail.co.uk/health/article-2467478/What-causes-dry-eye-syndrome-cure-treatment.html

MODEL DATA
MODEL
DEPLOYMENT
MODEL SCORING
MODEL TRAINING
Our challenge is…
…striving to be fast at everything

Most current, up-to-date data
Available as soon as it’s ready
Low latency at scale
FAST MODEL DATA

Distributed computing to crunch data fast
Elastic scaling with the public cloud
Speed from parallelism
FAST MODEL TRAINING

Model adapts to evolving customer landscape
Automatically refit the model and daily deploy
Seamlessly integrate with existing Java tech stack
FAST MODEL DEPLOYMENT

Response < 100 milliseconds
JVM-based model (i.e. POJO)
Predictive power vs. runtime complexity (speed)
Gradient boosting provided the best balance
FAST MODEL SCORING

Explore new technologies continuously
Ability to switch new models “on-the-fly”
Make the API faster
Incorporate new data sources
Resiliency, failover capabilities

Technology changes
Flexibility of the cloud
Keep it simple
Small empowered teams

What's hot

Vertical is the New Horizontal - MinneAnalytics 2016 Sri Ambati Keynote on AI

Sri Ambati

Strata 2016 - Architecting for Change: LinkedIn's new data ecosystem

Shirshanka Das

Shirshanka Das and Yael Garten describe how LinkedIn redesigned its data analytics ecosystem in the face of a significant product rewrite, covering the infrastructure changes that enable LinkedIn to roll out future product innovations with minimal downstream impact. Shirshanka and Yael explore the motivations and the building blocks for this reimagined data analytics ecosystem, the technical details of LinkedIn’s new client-side tracking infrastructure, its unified reporting platform, and its data virtualization layer on top of Hadoop and share lessons learned from data producers and consumers that are participating in this governance model. Along the way, they offer some anecdotal evidence during the rollout that validated some of their decisions and are also shaping the future roadmap of these efforts.

Productionizing H2O Models with Apache Spark

Sri Ambati

This meetup was recorded in Mountain View, CA on January 10th, 2019. Video recording from the meetup can be viewed here: https://youtu.be/yN26i7e_BtM Spark pipelines represent a powerful concept to support productionizing machine learning workflows. Their API allows to combine data processing with machine learning algorithms and opens opportunities for integration with various machine learning libraries. However, to benefit from the power of pipelines, their users need to have a freedom to choose and experiment with any machine learning algorithm or library. Therefore, we developed Sparkling Water that embeds H2O machine learning library of advanced algorithms into the Spark ecosystem and exposes them via pipeline API. Furthermore, the algorithms benefit from H2O MOJOs - Model Object Optimized - a powerful concept shared across entire H2O platform to store and exchange models. The MOJOs are designed for effective model deployment with focus on scoring speed, traceability, exchangeability, and backward compatibility. In this talk we will explain the architecture of Sparkling Water with focus on integration into the Spark pipelines and MOJOs. We’ll demonstrate creation of pipelines integrating H2O machine learning models and their deployments using Scala or Python. Furthermore, we will show how to utilize pre-trained model MOJOs with Spark pipelines. Speaker's Bio: Michal is the VP of Engineering at H2O.ai! Michal is a geek, developer, Java, Linux, programming languages enthusiast developing software for over 15 years. He obtained PhD from the Charles University in Prague in 2012 and post-doc at Purdue University. During his studies he was interested in construction of not only distributed but also embedded and real-time component-based systems using model-driven methods and domain-specific languages. He participated in design and development of various systems including SOFA and Fractal component systems or jPapabench control system.

An Architecture for Agile Machine Learning in Real-Time Applications

Johann Schleier-Smith

Presented at KDD, August 11, 2015. Abstract of the paper: Machine learning techniques have proved effective in recommender systems and other applications, yet teams working to deploy them lack many of the advantages that those in more established software disciplines today take for granted. The well-known Agile methodology advances projects in a chain of rapid development cycles, with subsequent steps often informed by production experiments. Support for such workflow in machine learning applications remains primitive. The platform developed at if(we) embodies a specific machine learning approach and a rigorous data architecture constraint, so allowing teams to work in rapid iterative cycles. We require models to consume data from a time-ordered event history, and we focus on facilitating creative feature engineering. We make it practical for data scientists to use the same model code in development and in production deployment, and make it practical for them to collaborate on complex models. We deliver real-time recommendations at scale, returning top results from among 10,000,000 candidates with sub-second response times and incorporating new updates in just a few seconds. Using the approach and architecture described here, our team can routinely go from ideas for new models to production-validated results within two weeks.

Nanda Vijaydev, BlueData - Deploying H2O in Large Scale Distributed Environme...

Sri Ambati

This session was recorded in San Francisco on February 5th, 2019 and can be viewed here: https://youtu.be/CgoxjmdyMiU This session will discuss how to get up and running quickly with containerized H2O environments (H2O Flow, Sparkling Water, and Driverless AI) at scale, in a multi-tenant architecture with a shared pool of resources using CPUs and/or GPUs. See how how you can spin up (and tear down) your H2O environments on-demand, with just a few mouse clicks. Find out how to enable quota management of GPU resources for greater efficiency, and easily connect your compute to your datasets for large-scale distributed machine learning. Learn how to operationalize your machine learning pipelines and deliver faster time-to-value for your AI initiative — while ensuring enterprise-grade security and high performance. Bio: Nanda Vijaydev is senior director of solutions at BlueData (now HPE) - where she leverages technologies like Hadoop, Spark, and TensorFlow to build solutions for enterprise analytics and machine learning use cases. Nanda has 10 years of experience in data management and data science. Previously, she worked on data science and big data projects in multiple industries, including healthcare and media; was a principal solutions architect at Silicon Valley Data Science; and served as director of solutions engineering at Karmasphere. Nanda has an in-depth understanding of the data analytics and data management space, particularly in the areas of data integration, ETL, warehousing, reporting, and machine learning.

Near realtime AI deployment with huge data and super low latency - Levi Brack...

Sri Ambati

Published on Nov 2, 2018 This talk was recorded in London on October 30th, 2018 and can be viewed here: https://youtu.be/erHt-1yBuUw Session: Travelport is a leading travel commerce platform that has truly huge data and many complex needs in terms of processing, performance and latency. This talk will demonstrate how we were able to harness big data technologies, H2O and cloud integration to deploy AI at scale and at low latency. The talk to cover practical advice taken from our AI journey; you will learn the successful strategies and the pitfalls of near real-time retraining ML models with streaming data and using all opensource technologies. Bio: As principal data scientist at Travelport, Levi Brackman leads a team of data scientists that are putting ML model into production. Prior to Travelport, Levi spent most of his career in the start-up world. He founded and led an organization that created innovative educational software applications and solutions used by high schools and youth organizations in the USA and Australia. Levi earned a PhD in the quantitative social sciences under the supervision of one the world's leading educational psychologists. He earned master’s degree from University College London and is author of a business book published in eight languages that was a bestseller in multiple countries. A native of North London (UK) Levi is married and has five children and now lives in Broomfield, Colorado.

H2O for Medicine and Intro to H2O in Python

Sri Ambati

Erin LeDell presents on machine learning for medicine using the H2O platform. She discusses how electronic health records, genomic data, medical images, and data from wearables can be used with machine learning for applications like predictive diagnostics, prognosis, and remote patient monitoring. H2O is an open source machine learning platform that provides algorithms like deep learning, random forests, and gradient boosting in an easy to use interface. It demonstrates an EEG example to predict eye state from brain signals.

Forget becoming a Data Scientist, become a Machine Learning Engineer instead

Data Con LA

Data Con LA 2020 Description Machine learning is an essential skill in today's job market. But when it comes to learning Machine Learning, beginners get lot of conflicting advice. I have been teaching ML for software engineers for years. In this talk *I will dis-spell some of the myths surrounding machine learning *give you solid, tangible plan on how to go about learning ML *and give you good pointers to start from *and steer you away from common mistakes Speaker Sujee Maniyam, Elephant Scale, Founder, Principal instructor

DevOps for DataScience

Stepan Pushkarev

Pm.ais ummit 180917 final

Nisha Talagala

How to Apply Machine Learning with R, H20, Apache Spark MLlib or PMML to Real...

Kai Wähner

This document provides an overview of how to apply big data analytics and machine learning to real-time processing. It discusses machine learning and big data analytics to analyze historical data and build models. These models can then be used in real-time processing without needing to be rebuilt, to take automated actions based on incoming data. The agenda includes sections on machine learning, analysis of historical data, real-time processing, and a live demo.

Driverless AI - Arno Candel, H2O.ai

Sri Ambati

This talk was given at H2O World 2018 NYC and can be viewed here: https://youtu.be/oxLZZMR1lVY Description: Driverless AI is H2O.ai's latest flagship product for automatic machine learning. It fully automates some of the most challenging and productive tasks in applied data science such as feature engineering, model tuning, model ensembling and model deployment. Driverless AI turns Kaggle-winning grandmaster recipes into production-ready code, and is specifically designed to avoid common mistakes such as under- or overfitting, data leakage or improper model validation, some of the hardest challenges in data science. Avoiding these pitfalls alone can save weeks or more for each model, and is necessary to achieve high modeling accuracy, especially for time-series problems. With Driverless AI, data scientists of all proficiency levels can train and deploy modeling pipelines with just a few clicks from the GUI. Advanced users can use the client API from Python. Driverless AI builds hundreds or thousands of models under the hood to select the best feature engineering and modeling pipeline for every specific problem such as churn prediction, fraud detection, real-estate pricing, store sales prediction, marketing ad campaigns and many more. To speed up training, Driverless AI uses highly optimized C++/CUDA algorithms to take full advantage of the latest compute hardware. For example, Driverless AI runs orders of magnitudes faster on the latest Nvidia GPU supercomputers on Intel and IBM platforms, both in the cloud or on premise. Driverless AI is fully supported on all major cloud providers. There are two more product innovations in Driverless AI: statistically rigorous automatic data visualization and machine learning interpretability with reason codes and explanations in plain English. Both help data scientists and analysts to quickly validate the data and the models. In this talk, we explain how Driverless AI works and show how easy it is to reach top 5% rankings for several highly competitive Kaggle competitions. (edited) Speaker's Bio: Arno Candel is the Chief Technology Officer at H2O.ai. He is the main committer of H2O-3 and Driverless AI and has been designing and implementing high-performance machine-learning algorithms since 2012. Previously, he spent a decade in supercomputing at ETH and SLAC and collaborated with CERN on next-generation particle accelerators. Arno holds a PhD and Masters summa cum laude in Physics from ETH Zurich, Switzerland. He was named “2014 Big Data All-Star” by Fortune Magazine and featured by ETH GLOBE in 2015. Follow him on Twitter: @ArnoCandel.

Microsoft AI Platform - AETHER Introduction

Karthik Murugesan

This document discusses Microsoft's investments and progress in AI. It covers: 1. How Microsoft has built an exabyte-scale data lake and AI tools to prepare data and build/train/deploy intelligent models at scale across the company. 2. Examples of how AI is being used across Microsoft businesses like Bing, Office, and healthcare to improve experiences and outcomes. 3. Microsoft's efforts to contribute to open standards like ONNX to promote interoperability and make AI more accessible to developers.

H2O World - Self Guiding Applications with Venkatesh Yadav

Sri Ambati

This document discusses self-guiding mobile applications that can provide engaging user experiences through predictive features. It presents a proof of concept for a photo editing app that uses crowdsourced usage data and machine learning algorithms to predict users' next actions and display relevant tools prominently. Performance results are promising. The document recommends exposing predictive models through REST APIs for loose coupling and continuous development, demonstrating how models trained on a platform can receive API requests and send predictions to applications.

ML-Ops: From Proof-of-Concept to Production Application

Hunter Carlisle

Ml ops past_present_future

Nisha Talagala

Operationalizing Machine Learning (Rajeev Dutt, CEO, Co-Founder, DimensionalM...

Amazon Web Services Korea

NeoPulse is a platform that aims to make AI ubiquitous by automating the creation, deployment, and management of AI models. It reduces the barriers to developing AI by requiring less code, having lower costs, and shorter project timelines compared to other platforms. The platform includes components like NeoPulse AI Studio, which can automate the creation of AI models, and NeoPulse Query Runtime, which allows applications to access models via an API. It supports a variety of data types and machine learning techniques. The document describes the end-to-end workflow on NeoPulse and provides examples of companies using it successfully.

Driverless AI - Intro + Interactive Hands-on Lab

Sri Ambati

Enjoy the webinar recording here: https://youtu.be/Lll1qwQJKVw. Driverless AI speeds up data science workflows by automating feature engineering, model tuning, ensembling, and model deployment. In this presentation, Arno Candel (CTO, H2O.ai), gives a quick overview and guide attendees through an interactive hands-on lab using Qwiklabs. Driverless AI turns Kaggle-winning recipes into production-ready code and is specifically designed to avoid common mistakes such as under or overfitting, data leakage or improper model validation. Avoiding these pitfalls alone can save weeks or more for each model, and is necessary to achieve high modeling accuracy. With Driverless AI, everyone can now train and deploy modeling pipelines with just a few clicks from the GUI. Advanced users can use the client/server API through a variety of languages such as Python, Java, C++, go, C# and many more. To speed up training, Driverless AI uses highly optimized C++/CUDA algorithms to take full advantage of the latest compute hardware. For example, Driverless AI runs orders of magnitudes faster on the latest Nvidia GPU supercomputers on Intel and IBM platforms, both in the cloud or on-premise. There are two more product innovations in Driverless AI: statistically rigorous automatic data visualization and interactive model interpretation with reason codes and explanations in plain English. Both help data scientists and analysts to quickly validate the data and models.

Introdution to Dataops and AIOps (or MLOps)

Adrien Blind

Scalable Automatic Machine Learning with H2O

Sri Ambati

In this presentation, Parul Pandey, will provide a history and overview of the field of “Automatic Machine Learning” (AutoML), followed by a detailed look inside H2O’s open source AutoML algorithm. H2O AutoML provides an easy-to-use interface which automates data pre-processing, training and tuning a large selection of candidate models (including multiple stacked ensemble models for superior model performance). The result of the AutoML run is a “leaderboard” of H2O models which can be easily exported for use in production. AutoML is available in all H2O interfaces (R, Python, Scala, web GUI) and due to the distributed nature of the H2O platform, can scale to very large datasets. The presentation will end with a demo of H2O AutoML in R and Python, including a handful of code examples to get you started using automatic machine learning on your own projects. Parul's Bio: Parul is a Data Science Evangelist here at H2O.ai. She combines Data Science, evangelism and community in her work. Her emphasis is to spread the information about H2O and Driverless AI to as many people as possible, She is also an active writer and has contributed towards various national and international publications.

What's hot (20)

Vertical is the New Horizontal - MinneAnalytics 2016 Sri Ambati Keynote on AI

Strata 2016 - Architecting for Change: LinkedIn's new data ecosystem

Productionizing H2O Models with Apache Spark

An Architecture for Agile Machine Learning in Real-Time Applications

Nanda Vijaydev, BlueData - Deploying H2O in Large Scale Distributed Environme...

Near realtime AI deployment with huge data and super low latency - Levi Brack...

H2O for Medicine and Intro to H2O in Python

Forget becoming a Data Scientist, become a Machine Learning Engineer instead

DevOps for DataScience

Pm.ais ummit 180917 final

How to Apply Machine Learning with R, H20, Apache Spark MLlib or PMML to Real...

Driverless AI - Arno Candel, H2O.ai

Microsoft AI Platform - AETHER Introduction

H2O World - Self Guiding Applications with Venkatesh Yadav

ML-Ops: From Proof-of-Concept to Production Application

Ml ops past_present_future

Operationalizing Machine Learning (Rajeev Dutt, CEO, Co-Founder, DimensionalM...

Driverless AI - Intro + Interactive Hands-on Lab

Introdution to Dataops and AIOps (or MLOps)

Scalable Automatic Machine Learning with H2O

Viewers also liked

Better Customer Experience with Data Science - Bernard Burg, Comcast

Sri Ambati

This document discusses using data science and machine learning to improve the customer experience at Comcast. It describes using error data from set-top boxes and customer behavior data to predict when customers will call and the reasons for their calls. H2O machine learning algorithms were able to more accurately predict calls and reasons compared to Spark ML, improving the customer service experience. Overall, adopting H2O's algorithms provided superior results, faster performance, and better use of memory compared to alternative tools.

H2O Advancements - Arno Candel

Sri Ambati

H2O has announced several new advancements including improved security features, semi-supervised deep learning, and capabilities for large Java models. New features were also added for tree algorithms like gradient boosted machines to improve accuracy, and H2O will now support convolutional and recurrent neural networks for tasks like image recognition and time series analysis. Deep Water was announced as H2O's next generation deep learning platform to bring enterprise deep learning capabilities.

Predicting Patient Outcomes in Real-Time at HCA

Sri Ambati

Data Scientist Allison Baker and Development Manager of Data Products Cody Hall work with a talented team of data scientists, software engineers, and web developers, and are building the framework and infrastructure to support a real-time prediction application, with the ability to scale across the entire company. Paramount to these efforts has been the capability of integrating the architecture for software production with the predictive models generated by H2O. This talk will review the processes by which HCA is building a pipeline to predict patient outcomes in real-time, heavily relying on H2O’s POJO scoring API and implemented in Clojure data processing. #h2ony - Powered by the open source machine learning software H2O.ai. Contributors welcome at: https://github.com/h2oai - To view videos on H2O open source machine learning software, go to: https://www.youtube.com/user/0xdata

Comcast Enterprise Network Services

vcardona

Visual Machine Learning - Tony Chu

Sri Ambati

Ads In A Quality Score World - SES NYC 2007

Jonathan Mendez

This document discusses how segmentation, targeting, and contextual relevance can improve ad performance and deliver ads that are relevant to users. It explains that segmentation involves dividing traffic into channels, campaigns, ad groups and keywords. Targeting then ensures the right ad copy, geography, and content is shown. Together segmentation and targeting aim to improve relevance and the user experience from first impression to conversion. Metrics like click-through rate and landing page relevance help search engines like Google determine ad quality and rank ads. The document also provides examples of how relevance can be improved for retirement ads and recipes through root keyword matching and multivariate testing of title, description and URL.

Display Ads Performance with Sugar

Tony Ford

Sugar aims to improve display ad performance and flexibility by decoupling ads from content. Initial attempts to fetch ads server-side and inject client-side failed due to issues with JavaScript execution and multiple document.write() calls. Using friendly iframes addressed cross-domain issues but ads still wanted to be URLs client-side. Tracking ad load times with Google Analytics provided insight but not control. The next step is a single HTTP request with creative injection for better performance than multiple URLs.

182 sspcc3 e_melkin

Society for Scholarly Publishing

The document proposes the Atypon Institution Registry, a single place for institutions to store authentication information like official names, contacts, IPs, and IDs. This would provide a centralized way for publishers, agents, and hosting vendors ("service providers") to access needed information from institutions. It would allow institutions to manage this information in one place rather than registering separately with each service provider. The registry would use standards like DOIs, XML, and COUNTER to organize and deliver data. An independent board would govern the registry, which would be available at no cost initially. Atypon has developed prototypes and seeks feedback on customer service and migrating existing institution metadata.

Search 4.0 Search Ads and Behavioral Targeting

Jonathan Mendez

The document discusses various techniques for targeted digital advertising and marketing, including search ads, behavioral targeting, personalization, source targeting, affinity targeting, temporal targeting, semantic targeting, and utilizing URL parameters and cookies. It provides examples of how these targeting methods can be used to increase metrics like click-through rate, conversion rate, and revenue per visit.

Targeting, Relevance, Engagement

atypon

Strata San Jose 2016: Scalable Ensemble Learning with H2O

Sri Ambati

This document discusses scalable ensemble learning using the H2O platform. It provides an overview of ensemble methods like bagging, boosting, and stacking. The stacking or Super Learner algorithm trains a "metalearner" to optimally combine the predictions from multiple "base learners". The H2O platform and its Ensemble package implement Super Learner and other ensemble methods for tasks like regression and classification. An R code demo is presented on training ensembles with H2O.

OnSite Tageting Strategy

Jonathan Mendez

This document outlines a marketing strategy that begins with understanding user intent and delivering relevant content through advanced technologies. It then discusses optimizing landing pages, sites, and applications to improve ROI. Finally, the strategy proposes building customer segments and affinities based on source and behavior data, hypothesizing relevance, testing segments, and monitoring high impact segments to personalize targeting.

It's Getting Personal: The Rise of Hyper-Targeted User Experiences - Colin Eagan

UXPA International

The User Experience community faces an all-new challenge in the rise of hyper-personalized web content, which allows sites to serve targeted messages to individual users based on behavioral or demographic data. While the technologies that support targeting grow more powerful by the day, the UX and content strategies behind them are often lacking -- or nonexistent. This has resulted in something we’ve all experienced online at one time or another: the creepy targeted ad. It’s time to fight back. While targeted marketing isn't going anywhere soon, the advent of Web Experience Management systems has democratized the landscape, meaning more clients have the ability to design personalized digital experiences. By re-framing the motivations behind targeting, UX designers have a significant opportunity to direct this power in ways that are user-centric, not marketer-centric. In this talk we’ll cover new tools and techniques to get started in the effort to take back personalized content for Good.

ISAX

Sri Ambati

ISAX is a time series data compression algorithm that can group similar patterns in billions of time series datasets. It is implemented on H2O's distributed architecture and can be used for clustering, classification, anomaly detection, and predictive analytics on compressed time series data from fields like IoT, finance, bioinformatics, and image/sound processing. Examples of ISAX code in H2O are provided.

Making Great Display Ads

Randy Oest II

This document discusses creating effective display ads. It begins by outlining various agency capabilities for qualitative and quantitative research as well as creative, media, and analytics. It then focuses on five key aspects of making great display ads: purpose, flow, design, production, and results. For each aspect, it provides tips and examples, such as common ad sizes, the importance of a clear call to action, testing different headlines, and examining conversion rates and click-through rates to evaluate results.

Driving In-Store Sales with Real-Time Personalization - Cyril Nigg, Catalina ...

Sri Ambati

Building a Real-Time Security Application Using Log Data and Machine Learning...

Sri Ambati

Stacked Ensembles in H2O

Sri Ambati

Ensemble machine learning methods are often used when the true prediction function is not easily approximated by a single algorithm. Practitioners may prefer ensemble algorithms when model performance is valued above other factors such as model complexity and training time. The Super Learner algorithm, also called "stacking", learns the optimal combination of the base learner fits. The latest version of H2O now contains a "Stacked Ensemble" method, which allows the user to stack H2O models into a Super Learner. The Stacked Ensemble method is the the native H2O version of stacking, previously only available in the h2oEnsemble R package, and now enables stacking from all the H2O APIs: Python, R, Scala, etc. Erin is a Statistician and Machine Learning Scientist at H2O.ai. Before joining H2O, she was the Principal Data Scientist at Wise.io (acquired by GE Digital) and Marvin Mobile Security (acquired by Veracode) and the founder of DataScientific, Inc. Erin received her Ph.D. from University of California, Berkeley. Her research focuses on ensemble machine learning, learning from imbalanced binary-outcome data, influence curve based variance estimation and statistical computing.

Capital One: Command center everywhere: Decentralizing the command center, pr...

SocialMedia.org

In their SocialMedia.org Member Meeting case study presentation, Capital One’s Principal Product Manager of Social Media Monitoring, Andrew Kauz, and Senior Brand Manager, Sherry Roper, explain how they decentralized their listening command center strategy to raise the importance of social data. They also talk about how their Spark Newsroom allows them to leverage social to offer additional service offerings to the team and business partners.

Capital One - Fairbank

Rohit Dhoundiyal

Mr. Richard Fairbank, CEO of Capital One Financial Corporation, believes that passion is important for successful leadership. He started his own business after graduating from Stanford business school, focusing on offering a variety of credit cards and allowing customers to transfer balances. As a leader, he emphasizes having commitment, courage, actively listening to others, maintaining a positive attitude, enjoying problem solving, and following your passion.

Viewers also liked (20)

Better Customer Experience with Data Science - Bernard Burg, Comcast

H2O Advancements - Arno Candel

Predicting Patient Outcomes in Real-Time at HCA

Comcast Enterprise Network Services

Visual Machine Learning - Tony Chu

Ads In A Quality Score World - SES NYC 2007

Display Ads Performance with Sugar

182 sspcc3 e_melkin

Search 4.0 Search Ads and Behavioral Targeting

Targeting, Relevance, Engagement

Strata San Jose 2016: Scalable Ensemble Learning with H2O

OnSite Tageting Strategy

It's Getting Personal: The Rise of Hyper-Targeted User Experiences - Colin Eagan

ISAX

Making Great Display Ads

Driving In-Store Sales with Real-Time Personalization - Cyril Nigg, Catalina ...

Building a Real-Time Security Application Using Log Data and Machine Learning...

Stacked Ensembles in H2O

Capital One: Command center everywhere: Decentralizing the command center, pr...

Capital One - Fairbank

Similar to Building Real Time Targeting Capabilities - Ryan Zotti, Subbu Thiruppathy - Capital One

Moving existing apps to the cloud

Ram Maddali

This document discusses moving existing applications to the cloud. It outlines different patterns for modernizing workloads including lift and shift, connect and extend, and rip and rewrite. It then describes a hands-on lab where participants will use the Red Hat Application Migration Toolkit to migrate a Java EE application from Weblogic to JBoss EAP and deploy it on OpenShift as a containerized "fast moving monolith". The goal is to learn how to modernize existing applications and deploy them in the cloud.

Cloud-Native Modernization or Death? A false dichotomy. | DevNation Tech Talk

Red Hat Developers

Are you familiar with the tight coupling of apps with their underlying platform that makes change hard or lack of scalability, performance, and flexibility of existing apps built with legacy technology or the fact that re-architecting apps cannot be done overnight? If yes to any of these, you probably think that you have “Cloud-Native Modernization or Death”. But what if there is another way that shows you the incremental steps to refactor the application to microservices and make use of Kubernetes/OpenShift to effectively deploy and manage it at scale on the cloud? This session guides developers on how to get started on their cloud-native journey, starting with monolithic application migration to a modern container platform using Kubernetes/OpenShift, and modernizing applications using microservices and Red Hat Cloud-Native Runtimes (Spring Boot and Quarkus).

Not Only Streams for Akademia JLabs

Konrad Malawski

This document summarizes Konrad Malawski's talk on reactive programming and related topics. Konrad discusses reactive streams, Akka toolkit for building concurrent applications, actors model for concurrency, and how circuit breakers can be used as a substitute for flow control. He also talks about the origins and development of reactive streams specification to provide a common set of semantics for backpressure.

NewFormatResume

Dileep Maurya

This document contains the resume of Dileep Kumar Maurya, an analytical full stack developer with over 6 years of experience in Java and web technologies. He has worked as a software engineer and senior software engineer on various projects for companies like Emaratech, Aricent Technology, HCL Technology, and Trila Soft Solution. His skills include Java, Spring, HTML, JavaScript, Angular, databases like Oracle, and version control tools like Git and SVN. He has experience in Agile methodologies and developing web applications using technologies like Java, Spring MVC, REST APIs, and front-end frameworks. He holds Java and Oracle certifications and received his B.Tech degree from Dr. K.N.

Why Startups Are Still On AWS

Amazon Web Services

This document discusses why startups continue to use Amazon Web Services (AWS). It notes that startups have limited resources and need to validate ideas quickly and at low cost. AWS allows startups to focus on building their product rather than infrastructure. It emphasizes lean startup principles of releasing minimal viable products, collecting user feedback, and iterating quickly. AWS services like EC2, S3, CloudFormation, and Elastic MapReduce help startups scale infrastructure as needed without large upfront costs. This allows startups to experiment more and innovate faster. Case studies show how startups have used AWS to handle growth, process large amounts of data, and deploy code continuously while maintaining flexibility.

Zagat.com Case Study (DrupalCon Denver 2012)

Phase2

This document discusses Zagat.com's migration from a legacy ASP.NET website to Drupal. It outlines problems with the old site such as being difficult to update and test. Solutions included using Drupal and related modules, optimizing front-end performance, automating deployments, and adding tools like Varnish, Jenkins, and Selenium. Non-Drupal talent was brought in to help with tasks like testing. Analytics and user testing were used to track goals and iterate quickly.

Agile integration workshop Seattle

Judy Breedlove

DevOps in Practice: When does "Practice" Become "Doing"?

Michael Elder

DevOps has emerged as the hot trend in development buzzword-ology. With a few quick paragraphs, it proposes to decimate all of the traditional problems you've encountered during your development experience. In IBM UrbanCode, we build products to help customers follow good DevOps practices. You may think DevOps is about the release process, but really it's about applying a mix of automation and operational practices earlier in your development life cycle so that rolling out to production becomes easier. DevOps promotes a focus on small-batch changes over large complex updates which are harder to predict and harder to roll back when problems occur. With greater velocity, rolling out smaller changes becomes more common place. Additionally, IBM UrbanCode makes extensive application of cloud technology that intercepts well with practices in DevOps around production-like environments. In this talk, Michael Elder describes how we practice DevOps internally with a mixture of IBM-built and open source tools. He'll discuss the areas that we do well and the challenges that we have with changing our culture around areas like test automation. On top of that, he'll describe how you can leverage these approaches in your own development process!

Cloud design pattern using azure

Karthikeyan VK

The document discusses various cloud design patterns and principles including CQRS, strangler, pipes and filters, circuit breaker, compensating transactions, static content hosting, gatekeeper, event sourcing, sharding, sidecar, and anti-patterns like busy database, no caching, and chatty I/O. It provides descriptions and examples of when to use each pattern to improve scalability, availability, and resilience of applications in the cloud.

Google App Engine for Java v0.0.2

Matthew McCullough

Google App Engine is a PaaS that allows developers to build and host web applications in the Google cloud. The document summarizes a workshop on using the Java runtime environment on GAE. It discusses the SDKs, deploying and managing apps on GAE, data storage using the datastore, and limitations like the 30-second request limit. The biggest benefits of GAE are scalability and low startup costs, while the hardest limit is the 30-second request processing time.

Choose Your Own Adventure with JHipster & Kubernetes - Denver JUG 2020

Matt Raible

Remember the choose your own adventure books that you used to read as a kid? This session is a reincarnation of a choose your own adventure book as a conference talk! You'll learn about Spring Boot, Docker, and Kubernetes in this talk, along with the choices you make in the following areas: * What kind of application architecture to build? Monolith or microservices? * Would you like to use Java or Kotlin? * MySQL, PostgreSQL, or MongoDB? * Spring MVC or Spring WebFlux? * Angular, React, or Vue.js? * PWA or mobile app? * Istio with Kubernetes or Kubernetes without Istio? GitHub repos of demos: * Monolith: https://github.com/mraible/healthy-hipster * Microservices: https://github.com/mraible/djug-microservices

App Mod 03: Monoliths to microservices with java ee and spring boot

Judy Breedlove

This document discusses transforming monolithic applications into microservices. It provides an overview of the benefits of breaking applications into smaller, autonomous services. The lab guides participants through implementing microservices from an existing monolith application using Java technologies like WildFly Swarm and Spring Boot on OpenShift. The lab uses the "strangler" pattern to gradually migrate functionality out of the monolith. The end result demonstrates how selected services from the monolith can be developed as independent microservices with their own databases.

All Change how the economics of Cloud will make you think differently about Java

Steve Poole

This document discusses how the economics of cloud computing will change how Java applications are developed. Cloud providers charge for computing resources on an hourly basis (e.g. $ per GB per hour), which means applications need to use resources efficiently. Java applications generally use more memory and have longer startup times than other languages. To be cost effective in the cloud, Java applications will need to reduce their memory footprint, decrease startup times, and be designed to fail and recover gracefully. The rise of APIs and microservices also requires changes to make Java more modular and efficient in constrained environments.

Continuous Deployment: The Dirty Details

Mike Brittain

JavaOne 2015 Keynote Presentation

ibmwebspheresoftware

The document discusses IBM's 20 year history with Java and innovations in Java performance. It highlights how IBM has optimized Java for big data and cloud computing. It also discusses how polyglot programming with languages like JavaScript and Node.js will be important for hybrid cloud architectures and driving developer productivity. The presentation concludes with a demo of Watson analytics APIs and deploying applications to Bluemix from Eclipse.

Puppet on a string

Alain Geenrits

This document discusses DevOps automation using Puppet Enterprise and VMware solutions. It describes how Puppet Enterprise can be used to automate the provisioning of multi-node applications on VMware vCloud Automation Center. This allows for self-service provisioning and lifecycle management of applications across heterogeneous infrastructure according to policies and approvals. It also enables drift remediation to fix configurations that deviate from the blueprint.

Fronted development trends - past, present and the future

Harijs Deksnis

This document provides an overview of past, present, and future trends in frontend development. In the past, JavaScript grew from a non-essential web language to an essential part of the modern web. Currently, JavaScript frameworks like React and Vue have become popular, and new programming models around components and functional programming have emerged. Looking ahead, trends like GraphQL, Progressive Web Apps, serverless architectures, AI/chatbots, and Web VR may shape the future of frontend development.

Beyond the IOPS: Flash Storage Essentials for Performance & Uptimes

SolarWinds

The document discusses all-flash storage arrays and their advantages over traditional storage arrays. The agenda includes discussing all-flash storage array performance and uptime essentials, best practices, and the importance of threshold and alerting. Traditional storage arrays have bottlenecks due to long seek times for hard disks, while all-flash arrays optimize CPU usage through faster solid state drives. All-flash arrays also offer benefits like data reduction, non-disruptive operations, and intelligent alerting based on key performance indicators. The presentation aims to demonstrate how all-flash arrays can provide simplicity, speed and scalability for applications.

Moving existing apps to the cloud

Tiera Fann, MBA

This document discusses migrating existing applications to the cloud and developing new applications. It covers moving monolithic applications to a microservices architecture using containers and orchestration. Approaches covered include lift and shift of monoliths, connecting and extending existing applications, and rewriting applications. Performance comparisons of various Java application runtimes are also presented.

Big Data And HTML5 (DevCon TLV 2012)

Ido Green

Similar to Building Real Time Targeting Capabilities - Ryan Zotti, Subbu Thiruppathy - Capital One (20)

Moving existing apps to the cloud

Cloud-Native Modernization or Death? A false dichotomy. | DevNation Tech Talk

Not Only Streams for Akademia JLabs

NewFormatResume

Why Startups Are Still On AWS

Zagat.com Case Study (DrupalCon Denver 2012)

Agile integration workshop Seattle

DevOps in Practice: When does "Practice" Become "Doing"?

Cloud design pattern using azure

Google App Engine for Java v0.0.2

Choose Your Own Adventure with JHipster & Kubernetes - Denver JUG 2020

App Mod 03: Monoliths to microservices with java ee and spring boot

All Change how the economics of Cloud will make you think differently about Java

Continuous Deployment: The Dirty Details

JavaOne 2015 Keynote Presentation

Puppet on a string

Fronted development trends - past, present and the future

Beyond the IOPS: Flash Storage Essentials for Performance & Uptimes

Moving existing apps to the cloud

Big Data And HTML5 (DevCon TLV 2012)

More from Sri Ambati

GenAISummit 2024 May 28 Sri Ambati Keynote: AGI Belongs to The Community in O...

Sri Ambati

H2O.ai CEO/Founder: Sri Ambati Keynote at Wells Fargo Day

Sri Ambati

This document provides an overview of H2O.ai, an AI company that offers products and services to democratize AI. It mentions that H2O products are backed by 10% of the world's top data scientists from Kaggle and that H2O has customers in 7 of the top 10 banks, 4 of the top 10 insurance companies, and top manufacturing companies. It also provides details on H2O's founders, funding, customers, products, and vision to make AI accessible to more organizations.

Generative AI Masterclass - Model Risk Management.pptx

Sri Ambati

Here are some key points about benchmarking and evaluating generative AI models like large language models: - Foundation models require large, diverse datasets to be trained on in order to learn broad language skills and knowledge. Fine-tuning can then improve performance on specific tasks. - Popular benchmarks evaluate models on tasks involving things like commonsense reasoning, mathematics, science questions, generating truthful vs false responses, and more. This helps identify model capabilities and limitations. - Custom benchmarks can also be designed using tools like Eval Studio to systematically test models on specific applications or scenarios. Both automated and human evaluations are important. - Leaderboards like HELM aggregate benchmark results to compare how different models perform across a wide range of tests and metrics.

AI and the Future of Software Development: A Sneak Peek

Sri Ambati

LLMOps: Match report from the top of the 5th

Sri Ambati

The document discusses LLMOps (Large Language Model Operations) compared to traditional MLOps. Some key points: - LLMOps and MLOps face similar challenges across the development lifecycle, but LLMOps requires more GPU resources and integration is faster due to more models in each application. Evaluation is also less clear. - The LLMOps field is around the 5th generation of models, with debates around proprietary vs open source models, and balancing privacy, cost and control. - LLMOps platforms are emerging to provide solutions for tasks like prompting, embedding databases, evaluation, and governance, similar to how MLOps platforms have evolved.

Building, Evaluating, and Optimizing your RAG App for Production

Sri Ambati

The document discusses optimizing question answering systems called RAG (Retrieve-and-Generate) stacks. It outlines challenges with naive RAG approaches and proposes solutions like improved data representations, advanced retrieval techniques, and fine-tuning large language models. Table stakes optimizations include tuning chunk sizes, prompt engineering, and customizing LLMs. More advanced techniques involve small-to-big retrieval, multi-document agents, embedding fine-tuning, and LLM fine-tuning.

Building LLM Solutions using Open Source and Closed Source Solutions in Coher...

Sri Ambati

Sandeep Singh, Head of Applied AI Computer Vision, Beans.ai H2O Open Source GenAI World SF 2023 In the modern era of machine learning, leveraging both open-source and closed-source solutions has become paramount for achieving cutting-edge results. This talk delves into the intricacies of seamlessly integrating open-source Large Language Model (LLM) solutions like Vicuna, Falcon, and Llama with industry giants such as ChatGPT and Google's Palm. As the demand for fine-tuned and specialized datasets grows, it is imperative to understand the synergy between these tools. Attendees will gain insights into best practices for building and enriching datasets tailored for fine-tuning tasks, ensuring that their LLM projects are both robust and efficient. Through real-world examples and hands-on demonstrations, this talk will equip attendees with the knowledge to harness the power of both open and closed-source tools in a coherent and effective manner.

Risk Management for LLMs

Sri Ambati

Patrick Hall, Professor, AI Risk Management, The George Washington University H2O Open Source GenAI World SF 2023 Language models are incredible engineering breakthroughs but require auditing and risk management before productization. These systems raise concerns about toxicity, transparency and reproducibility, intellectual property licensing and ownership, disinformation and misinformation, supply chains, and more. How can your organization leverage these new tools without taking on undue or unknown risks? While language models and associated risk management are in their infancy, a small number of best practices in governance and risk are starting to emerge. If you have a language model use case in mind, want to understand your risks, and do something about them, this presentation is for you!

Open-Source AI: Community is the Way

Sri Ambati

Dr. Alexy Khrabrov, Open Source Science Community Director, IBM H2O Open Source GenAI World SF 2023 In this talk, Dr. Alexy Khrabrov, recently elected Chair of the new Generative AI Commons at Linux Foundation for AI & Data, outlines the OSS AI landscape, challenges, and opportunities. With new models and frameworks being unveiled weekly, one thing remains constant: community building and validation of all aspects of AI is key to reliable and responsible AI we can use for business and society needs. Industrial AI is one key area where such community validation can prove invaluable.

Building Custom GenAI Apps at H2O

Sri Ambati

The document announces the launch of the H2O GenAI App Store, which provides a collection of applications that make it easier for average users to leverage large language models through custom interfaces for specific tasks like getting gardening advice or feedback on code. The app store is designed to accelerate the development of these GenAI apps using the H2O Wave platform and provides access to H2OGPTE for retrieval augmented generation and language model calls. Developers can also contribute their own apps through the GitHub repository listed.

Applied Gen AI for the Finance Vertical

Sri Ambati

Megan Kurka, Vice President, Customer Data Scientist, H2O.ai H2O Open Source GenAI World SF 2023 Discover the transformative power of Applied Gen AI. Learn how the H2O team builds customized applications and workflows that integrate capabilities of Gen AI and AutoML specifically designed to address and enhance financial use cases. Explore real world examples, learn best practices, and witness firsthand how our innovative solutions are reshaping the landscape of finance technology.

Cutting Edge Tricks from LLM Papers

Sri Ambati

This document discusses techniques for improving language models (LLMs) discussed in recent papers. It describes building blocks of LLMs like fine-tuning, foundation training, memory, and databases. Specific techniques covered include LIMA which uses 1,000 carefully curated examples, instruction backtranslation to generate question-answer pairs, fine-tuning models on API examples like Gorilla, and reducing false answers through techniques like not agreeing with incorrect user opinions. The goal is to discuss cutting edge tricks to build better LLMs.

Practitioner's Guide to LLMs: Exploring Use Cases and a Glimpse Beyond Curren...

Sri Ambati

Pascal Pfeiffer, Principal Data Scientist, H2O.ai H2O Open Source GenAI World SF 2023 This talk dives into the expansive ecosystem of Large Language Models (LLMs), offering practitioners an insightful guide to various relevant applications, from natural language understanding to creative content generation. While exploring use cases across different industries, it also honestly addresses the current limitations of LLMs and anticipates future advancements.

Open Source h2oGPT with Retrieval Augmented Generation (RAG), Web Search, and...

Sri Ambati

KGM Mastering Classification and Regression with LLMs: Insights from Kaggle C...

Sri Ambati

This document discusses using large language models (LLMs) for text classification tasks. It begins by describing how LLMs are commonly used for text generation and question answering. For classification, models are usually trained supervised on labeled data. The document then explores using LLMs for zero-shot classification without training, and techniques like fine-tuning LLMs on tasks to improve performance. It provides an example of fine-tuning an LLM on a financial sentiment dataset. The document concludes by describing H2O.ai's LLM Studio tool for fine-tuning and a few Kaggle competitions where LLMs achieved success in text classification.

LLM Interpretability

Sri Ambati

1) Generative AI (GenAI) enables the creation of novel content by learning patterns in unstructured data rather than labeling outputs like traditional AI. 2) Both traditional and generative AI models lack transparency and may contain biases, but generative models can additionally hallucinate or leak private information. 3) To interpret generative models, researchers evaluate accuracy globally by checking for hallucinations or undesirable content, and locally by confirming the quality of individual responses.

Never Reply to an Email Again

Sri Ambati

From Rapid Prototypes to an end-to-end Model Deployment: an AI Hedge Fund Use...

Sri Ambati

Numerai is an open, crowd-sourced hedge fund powered by predictions from data scientists around the world. In return, participants are rewarded with weekly payouts in crypto. In this talk, Joe will give an overview of the Numerai tournament based on his own experience. He will then explain how he automates the time-consuming tasks such as testing different modelling strategies, scoring new datasets, submitting predictions to Numerai as well as monitoring model performance with H2O Driverless AI and R.

AI Foundations Course Module 1 - Shifting to the Next Step in Your AI Transfo...

Sri Ambati

In this session, you will learn about what you should do after you’ve taken an AI transformation baseline. Over the span of this session, we will discuss the next steps in moving toward AI readiness through alignment of talent and tools to drive successful adoption and continuous use within an organization. To find additional videos on AI courses, earn badges, join the courses at H2O.ai Learning Center: https://training.h2o.ai/products/ai-foundations-course To find the Youtube video about this presentation: https://youtu.be/K1Cl3x3rd8g Speaker: Chemere Davis (H2O.ai - Senior Data Scientist Training Specialist)

AI Foundations Course Module 1 - An AI Transformation Journey

Sri Ambati

The chances of successfully implementing AI strategies within an organization significantly improve when you can recognize where your organization is on the maturity scale. Over this course, you will learn the keys to unlocking value with AI which include asking the right questions about the problems you are solving and ensuring you have the right cross-section of talent, tools, and resources. By the end of this module, you should be able to recognize where your organization is on the AI transformation spectrum and identify some strategies that can get you to the next stage in your journey. To find additional videos on AI courses, earn badges, join the courses at H2O.ai Learning Center: https://training.h2o.ai/products/ai-foundations-course To find the Youtube video about this presentation: https://youtu.be/PJgr2epM6qs Speakers: Chemere Davis (H2O.ai - Senior Data Scientist Training Specialist) Ingrid Burton (H2O.ai - CMO)

More from Sri Ambati (20)

GenAISummit 2024 May 28 Sri Ambati Keynote: AGI Belongs to The Community in O...

H2O.ai CEO/Founder: Sri Ambati Keynote at Wells Fargo Day

Generative AI Masterclass - Model Risk Management.pptx

AI and the Future of Software Development: A Sneak Peek

LLMOps: Match report from the top of the 5th

Building, Evaluating, and Optimizing your RAG App for Production

Building LLM Solutions using Open Source and Closed Source Solutions in Coher...

Risk Management for LLMs

Open-Source AI: Community is the Way

Building Custom GenAI Apps at H2O

Applied Gen AI for the Finance Vertical

Cutting Edge Tricks from LLM Papers

Practitioner's Guide to LLMs: Exploring Use Cases and a Glimpse Beyond Curren...

Open Source h2oGPT with Retrieval Augmented Generation (RAG), Web Search, and...

KGM Mastering Classification and Regression with LLMs: Insights from Kaggle C...

LLM Interpretability

Never Reply to an Email Again

From Rapid Prototypes to an end-to-end Model Deployment: an AI Hedge Fund Use...

AI Foundations Course Module 1 - Shifting to the Next Step in Your AI Transfo...

AI Foundations Course Module 1 - An AI Transformation Journey

Recently uploaded

Streamlining Legacy Complexity Through Modernization

sanjay singh

Malviya Nagar @ℂall @Girls ꧁❤ 9873940964 ❤꧂VIP Jina Singh Top Model Safe

butwhat24

BIGPPTTTTTTTTtttttttttttttttttttttt.pptx

RajdeepPaul47

[D3T2S03] Data&AI Roadshow 2024 - Amazon DocumentDB 실습

Amazon Web Services Korea

Amazon DocumentDB(MongoDB와 호환됨)는 빠르고 안정적이며 완전 관리형 데이터베이스 서비스입니다. Amazon DocumentDB를 사용하면 클라우드에서 MongoDB 호환 데이터베이스를 쉽게 설치, 운영 및 규모를 조정할 수 있습니다. Amazon DocumentDB를 사용하면 MongoDB에서 사용하는 것과 동일한 애플리케이션 코드를 실행하고 동일한 드라이버와 도구를 사용하는 것을 실습합니다.

[D3T1S04] Aurora PostgreSQL performance monitoring and troubleshooting by use...

Amazon Web Services Korea

NPS_Presentation_V3.pptx it is regarding National pension scheme

ASISHSABAT3

Australian Catholic University degree offer diploma Transcript

taqyea

学历认证补办制【微信：A575476】【(ACU毕业证）澳大利亚天主教大学毕业证成绩单offer】【微信：A575476】（留信学历认证永久存档查询）采用学校原版纸张，特殊工艺完全按照原版一比一制作（包括：隐形水印，阴影底纹，钢印LOGO烫金烫银，LOGO烫金烫银复合重叠，文字图案浮雕，激光镭射，紫外荧光，温感，复印防伪）行业标杆！精益求精，诚心合作，真诚制作！多年品质 ,按需精细制作，24小时接单,全套进口原装设备，十五年致力于帮助留学生解决难题，业务范围有加拿大、英国、澳洲、韩国、美国、新加坡，新西兰等学历材料，包您满意。【业务选择办理准则】一、工作未确定，回国需先给父母、亲戚朋友看下文凭的情况，办理一份就读学校的毕业证【微信：A575476】文凭即可二、回国进私企、外企、自己做生意的情况，这些单位是不查询毕业证真伪的，而且国内没有渠道去查询国外文凭的真假，也不需要提供真实教育部认证。鉴于此，办理一份毕业证【微信：A575476】即可三、进国企，银行，事业单位，考公务员等等，这些单位是必需要提供真实教育部认证的，办理教育部认证所需资料众多且烦琐，所有材料您都必须提供原件，我们凭借丰富的经验，快捷的绿色通道帮您快速整合材料，让您少走弯路。留信网认证的作用: 1:该专业认证可证明留学生真实身份【微信：A575476】 2:同时对留学生所学专业登记给予评定 3:国家专业人才认证中心颁发入库证书 4:这个认证书并且可以归档倒地方 5:凡事获得留信网入网的信息将会逐步更新到个人身份内，将在公安局网内查询个人身份证信息后，同步读取人才网入库信息 6:个人职称评审加20分 7:个人信誉贷款加10分 8:在国家人才网主办的国家网络招聘大会中纳入资料，供国家高端企业选择人才 → 【关于价格问题（保证一手价格）我们所定的价格是非常合理的，而且我们现在做得单子大多数都是代理和回头客户介绍的所以一般现在有新的单子我给客户的都是第一手的代理价格，因为我想坦诚对待大家不想跟大家在价格方面浪费时间对于老客户或者被老客户介绍过来的朋友，我们都会适当给一些优惠。选择实体注册公司办理，更放心，更安全！我们的承诺：可来公司面谈，可签订合同，会陪同客户一起到教育部认证窗口递交认证材料，客户在教育部官方认证查询网站查询到认证通过结果后付款，不成功不收费！办理(ACU毕业证）澳大利亚天主教大学毕业证【微信：A575476】外观非常精致，由特殊纸质材料制成，上面印有校徽、校名、毕业生姓名、专业等信息。办理(ACU毕业证）澳大利亚天主教大学毕业证【微信：A575476】格式相对统一，各专业都有相应的模板。通常包括以下部分：校徽：象征着学校的荣誉和传承。校名:学校英文全称授予学位：本部分将注明获得的具体学位名称。毕业生姓名：这是最重要的信息之一，标志着该证书是由特定人员获得的。颁发日期：这是毕业正式生效的时间，也代表着毕业生学业的结束。其他信息：根据不同的专业和学位，可能会有一些特定的信息或章节。办理(ACU毕业证）澳大利亚天主教大学毕业证【微信：A575476】价值很高，需要妥善保管。一般来说，应放置在安全、干燥、防潮的地方，避免长时间暴露在阳光下。如需使用，最好使用复印件而不是原件，以免丢失。综上所述，办理(ACU毕业证）澳大利亚天主教大学毕业证【微信：A575476 】是证明身份和学历的高价值文件。外观简单庄重，格式统一，包括重要的个人信息和发布日期。对持有人来说，妥善保管是非常重要的。

EGU2020-10385_presentation LSTM algorithm

fatimaezzahraboumaiz2

MUMBAI MONTHLY RAINFALL CAPSTONE PROJECT

GaneshGanesh399816

### Data Description and Analysis Summary for Presentation #### 1. **Importing Libraries** Libraries used: - `pandas`, `numpy`: Data manipulation - `matplotlib`, `seaborn`: Data visualization - `scikit-learn`: Machine learning utilities - `statsmodels`, `pmdarima`: Statistical modeling - `keras`: Deep learning models #### 2. **Loading and Exploring the Dataset** **Dataset Overview:** - **Source:** CSV file (`mumbai-monthly-rains.csv`) - **Columns:** - `Year`: The year of the recorded data. - `Jan` to `Dec`: Monthly rainfall data. - `Total`: Total annual rainfall. **Initial Data Checks:** - Displayed first few rows. - Summary statistics (mean, standard deviation, min, max). - Checked for missing values. - Verified data types. **Visualizations:** - **Annual Rainfall Time Series:** Trends in annual rainfall over the years. - **Monthly Rainfall Over Years:** Patterns and variations in monthly rainfall. - **Yearly Total Rainfall Distribution:** Distribution and frequency of annual rainfall. - **Box Plots for Monthly Data:** Spread and outliers in monthly rainfall. - **Correlation Matrix of Monthly Rainfall:** Relationships between different months' rainfall. #### 3. **Data Transformation** **Steps:** - Ensured 'Year' column is of integer type. - Created a datetime index. - Converted monthly data to a time series format. - Created lag features to capture past values. - Generated rolling statistics (mean, standard deviation) for different window sizes. - Added seasonal indicators (dummy variables for months). - Dropped rows with NaN values. **Result:** - Transformed dataset with additional features ready for time series analysis. #### 4. **Data Splitting** **Procedure:** - Split the data into features (`X`) and target (`y`). - Further split into training (80%) and testing (20%) sets without shuffling to preserve time series order. **Result:** - Training set: `(X_train, y_train)` - Testing set: `(X_test, y_test)` #### 5. **Automated Hyperparameter Tuning** **Tool Used:** `pmdarima` - Automatically selected the best parameters for the SARIMA model. - Evaluated using metrics such as AIC and BIC. **Output:** - Best SARIMA model parameters and statistical summary. #### 6. **SARIMA Model** **Steps:** - Fit the SARIMA model using the training data. - Evaluated on both training and testing sets using MAE and RMSE. **Output:** - **Train MAE:** Indicates accuracy on training data. - **Test MAE:** Indicates accuracy on unseen data. - **Train RMSE:** Measures average error magnitude on training data. - **Test RMSE:** Measures average error magnitude on testing data. #### 7. **LSTM Model** **Preparation:** - Reshaped data for LSTM input. - Converted data to `float32`. **Model Building and Training:** - Built an LSTM model with one LSTM layer and one Dense layer. - Trained the model on the training data. **Evaluation:** - Evaluated on both training and testing sets using MAE and RMSE. **Output:** - **Train MAE:** Accuracy on training data. - **T

University of Toronto degree offer diploma Transcript

taqyea

学历认证补办制【微信：A575476】【(UofT毕业证）多伦多大学毕业证成绩单offer】【微信：A575476】（留信学历认证永久存档查询）采用学校原版纸张，特殊工艺完全按照原版一比一制作（包括：隐形水印，阴影底纹，钢印LOGO烫金烫银，LOGO烫金烫银复合重叠，文字图案浮雕，激光镭射，紫外荧光，温感，复印防伪）行业标杆！精益求精，诚心合作，真诚制作！多年品质 ,按需精细制作，24小时接单,全套进口原装设备，十五年致力于帮助留学生解决难题，业务范围有加拿大、英国、澳洲、韩国、美国、新加坡，新西兰等学历材料，包您满意。【业务选择办理准则】一、工作未确定，回国需先给父母、亲戚朋友看下文凭的情况，办理一份就读学校的毕业证【微信：A575476】文凭即可二、回国进私企、外企、自己做生意的情况，这些单位是不查询毕业证真伪的，而且国内没有渠道去查询国外文凭的真假，也不需要提供真实教育部认证。鉴于此，办理一份毕业证【微信：A575476】即可三、进国企，银行，事业单位，考公务员等等，这些单位是必需要提供真实教育部认证的，办理教育部认证所需资料众多且烦琐，所有材料您都必须提供原件，我们凭借丰富的经验，快捷的绿色通道帮您快速整合材料，让您少走弯路。留信网认证的作用: 1:该专业认证可证明留学生真实身份【微信：A575476】 2:同时对留学生所学专业登记给予评定 3:国家专业人才认证中心颁发入库证书 4:这个认证书并且可以归档倒地方 5:凡事获得留信网入网的信息将会逐步更新到个人身份内，将在公安局网内查询个人身份证信息后，同步读取人才网入库信息 6:个人职称评审加20分 7:个人信誉贷款加10分 8:在国家人才网主办的国家网络招聘大会中纳入资料，供国家高端企业选择人才 → 【关于价格问题（保证一手价格）我们所定的价格是非常合理的，而且我们现在做得单子大多数都是代理和回头客户介绍的所以一般现在有新的单子我给客户的都是第一手的代理价格，因为我想坦诚对待大家不想跟大家在价格方面浪费时间对于老客户或者被老客户介绍过来的朋友，我们都会适当给一些优惠。选择实体注册公司办理，更放心，更安全！我们的承诺：可来公司面谈，可签订合同，会陪同客户一起到教育部认证窗口递交认证材料，客户在教育部官方认证查询网站查询到认证通过结果后付款，不成功不收费！办理(UofT毕业证）多伦多大学毕业证【微信：A575476】外观非常精致，由特殊纸质材料制成，上面印有校徽、校名、毕业生姓名、专业等信息。办理(UofT毕业证）多伦多大学毕业证【微信：A575476】格式相对统一，各专业都有相应的模板。通常包括以下部分：校徽：象征着学校的荣誉和传承。校名:学校英文全称授予学位：本部分将注明获得的具体学位名称。毕业生姓名：这是最重要的信息之一，标志着该证书是由特定人员获得的。颁发日期：这是毕业正式生效的时间，也代表着毕业生学业的结束。其他信息：根据不同的专业和学位，可能会有一些特定的信息或章节。办理(UofT毕业证）多伦多大学毕业证【微信：A575476】价值很高，需要妥善保管。一般来说，应放置在安全、干燥、防潮的地方，避免长时间暴露在阳光下。如需使用，最好使用复印件而不是原件，以免丢失。综上所述，办理(UofT毕业证）多伦多大学毕业证【微信：A575476 】是证明身份和学历的高价值文件。外观简单庄重，格式统一，包括重要的个人信息和发布日期。对持有人来说，妥善保管是非常重要的。

From Clues to Connections: How Social Media Investigators Expose Hidden Networks

Milind Agarwal

Rohini @ℂall @Girls ꧁❤ 9873940964 ❤꧂VIP Megha Singla Top Model Safe

depikasharma

LLM powered Contract Compliance Application.pptx

Jyotishko Biswas

Sunshine Coast University diploma

cwavvyy

原版一模一样【微信：741003700 】【阳光海岸大学毕业证成绩单】【微信：741003700 】学位证，留信学历认证（真实可查，永久存档）原件一模一样纸张工艺/offer、在读证明、外壳等材料/诚信可靠,可直接看成品样本，帮您解决无法毕业带来的各种难题！外壳，原版制作，诚信可靠，可直接看成品样本。行业标杆！精益求精，诚心合作，真诚制作！多年品质 ,按需精细制作，24小时接单,全套进口原装设备。十五年致力于帮助留学生解决难题，包您满意。本公司拥有海外各大学样板无数，能完美还原。 1:1完美还原海外各大学毕业材料上的工艺：水印，阴影底纹，钢印LOGO烫金烫银，LOGO烫金烫银复合重叠。文字图案浮雕、激光镭射、紫外荧光、温感、复印防伪等防伪工艺。材料咨询办理、认证咨询办理请加学历顾问Q/微741003700 【主营项目】一.毕业证【q微741003700】成绩单、使馆认证、教育部认证、雅思托福成绩单、学生卡等！二.真实使馆公证(即留学回国人员证明,不成功不收费) 三.真实教育部学历学位认证（教育部存档！教育部留服网站永久可查）四.办理各国各大学文凭(一对一专业服务,可全程监控跟踪进度) 如果您处于以下几种情况： ◇在校期间，因各种原因未能顺利毕业……拿不到官方毕业证【q/微741003700】 ◇面对父母的压力，希望尽快拿到； ◇不清楚认证流程以及材料该如何准备； ◇回国时间很长，忘记办理； ◇回国马上就要找工作，办给用人单位看； ◇企事业单位必须要求办理的 ◇需要报考公务员、购买免税车、落转户口 ◇申请留学生创业基金留信网认证的作用: 1:该专业认证可证明留学生真实身份 2:同时对留学生所学专业登记给予评定 3:国家专业人才认证中心颁发入库证书 4:这个认证书并且可以归档倒地方 5:凡事获得留信网入网的信息将会逐步更新到个人身份内，将在公安局网内查询个人身份证信息后，同步读取人才网入库信息 6:个人职称评审加20分 7:个人信誉贷款加10分 8:在国家人才网主办的国家网络招聘大会中纳入资料，供国家高端企业选择人才办理阳光海岸大学毕业证【微信：741003700 】外观非常简单，由纸质材料制成，上面印有校徽、校名、毕业生姓名、专业等信息。办理阳光海岸大学毕业证【微信：741003700 】格式相对统一，各专业都有相应的模板。通常包括以下部分：校徽：象征着学校的荣誉和传承。校名:学校英文全称授予学位：本部分将注明获得的具体学位名称。毕业生姓名：这是最重要的信息之一，标志着该证书是由特定人员获得的。颁发日期：这是毕业正式生效的时间，也代表着毕业生学业的结束。其他信息：根据不同的专业和学位，可能会有一些特定的信息或章节。办理阳光海岸大学毕业证【微信：741003700 】价值很高，需要妥善保管。一般来说，应放置在安全、干燥、防潮的地方，避免长时间暴露在阳光下。如需使用，最好使用复印件而不是原件，以免丢失。综上所述，办理阳光海岸大学毕业证【微信：741003700 】是证明身份和学历的高价值文件。外观简单庄重，格式统一，包括重要的个人信息和发布日期。对持有人来说，妥善保管是非常重要的。

Seamlessly Pay Online, Pay In Stores or Send Money

gargtinna79

Noida Extension @ℂall @Girls ꧁❤ 9873940964 ❤꧂VIP Vishakha Singla Top Model Safe

kumkum tuteja$A17

iot paper presentation FINAL EDIT by kiran.pptx

KiranKumar139571

Niagara College degree offer diploma Transcript

taqyea

原版制作【微信：A575476】【(NC毕业证)尼亚加拉学院毕业证成绩单offer】【微信：A575476】（留信学历认证永久存档查询）采用学校原版纸张（包括：隐形水印，阴影底纹，钢印LOGO烫金烫银，LOGO烫金烫银复合重叠，文字图案浮雕，激光镭射，紫外荧光，温感，复印防伪）行业标杆！精益求精，诚心合作，真诚制作！多年品质 ,按需精细制作，24小时接单,全套进口原装设备，十五年致力于帮助留学生解决难题，业务范围有加拿大、英国、澳洲、韩国、美国、新加坡，新西兰等学历材料，包您满意。【业务选择办理准则】一、工作未确定，回国需先给父母、亲戚朋友看下文凭的情况，办理一份就读学校的毕业证【微信：A575476】文凭即可二、回国进私企、外企、自己做生意的情况，这些单位是不查询毕业证真伪的，而且国内没有渠道去查询国外文凭的真假，也不需要提供真实教育部认证。鉴于此，办理一份毕业证【微信：A575476】即可三、进国企，银行，事业单位，考公务员等等，这些单位是必需要提供真实教育部认证的，办理教育部认证所需资料众多且烦琐，所有材料您都必须提供原件，我们凭借丰富的经验，快捷的绿色通道帮您快速整合材料，让您少走弯路。留信网认证的作用: 1:该专业认证可证明留学生真实身份【微信：A575476】 2:同时对留学生所学专业登记给予评定 3:国家专业人才认证中心颁发入库证书 4:这个认证书并且可以归档倒地方 5:凡事获得留信网入网的信息将会逐步更新到个人身份内，将在公安局网内查询个人身份证信息后，同步读取人才网入库信息 6:个人职称评审加20分 7:个人信誉贷款加10分 8:在国家人才网主办的国家网络招聘大会中纳入资料，供国家高端企业选择人才 → 【关于价格问题（保证一手价格）我们所定的价格是非常合理的，而且我们现在做得单子大多数都是代理和回头客户介绍的所以一般现在有新的单子我给客户的都是第一手的代理价格，因为我想坦诚对待大家不想跟大家在价格方面浪费时间对于老客户或者被老客户介绍过来的朋友，我们都会适当给一些优惠。选择实体注册公司办理，更放心，更安全！我们的承诺：可来公司面谈，可签订合同，会陪同客户一起到教育部认证窗口递交认证材料，客户在教育部官方认证查询网站查询到认证通过结果后付款，不成功不收费！办理(NC毕业证)尼亚加拉学院毕业证【微信：A575476】外观非常精致，由特殊纸质材料制成，上面印有校徽、校名、毕业生姓名、专业等信息。办理(NC毕业证)尼亚加拉学院毕业证【微信：A575476】格式相对统一，各专业都有相应的模板。通常包括以下部分：校徽：象征着学校的荣誉和传承。校名:学校英文全称授予学位：本部分将注明获得的具体学位名称。毕业生姓名：这是最重要的信息之一，标志着该证书是由特定人员获得的。颁发日期：这是毕业正式生效的时间，也代表着毕业生学业的结束。其他信息：根据不同的专业和学位，可能会有一些特定的信息或章节。办理(NC毕业证)尼亚加拉学院毕业证【微信：A575476】价值很高，需要妥善保管。一般来说，应放置在安全、干燥、防潮的地方，避免长时间暴露在阳光下。如需使用，最好使用复印件而不是原件，以免丢失。综上所述，办理(NC毕业证)尼亚加拉学院毕业证【微信：A575476 】是证明身份和学历的高价值文件。外观简单庄重，格式统一，包括重要的个人信息和发布日期。对持有人来说，妥善保管是非常重要的。

University of the Sunshine Coast degree offer diploma Transcript

taqyea

学历认证补办制【微信：A575476】【(USC毕业证）阳光海岸大学毕业证成绩单offer】【微信：A575476】（留信学历认证永久存档查询）采用学校原版纸张，特殊工艺完全按照原版一比一制作（包括：隐形水印，阴影底纹，钢印LOGO烫金烫银，LOGO烫金烫银复合重叠，文字图案浮雕，激光镭射，紫外荧光，温感，复印防伪）行业标杆！精益求精，诚心合作，真诚制作！多年品质 ,按需精细制作，24小时接单,全套进口原装设备，十五年致力于帮助留学生解决难题，业务范围有加拿大、英国、澳洲、韩国、美国、新加坡，新西兰等学历材料，包您满意。【业务选择办理准则】一、工作未确定，回国需先给父母、亲戚朋友看下文凭的情况，办理一份就读学校的毕业证【微信：A575476】文凭即可二、回国进私企、外企、自己做生意的情况，这些单位是不查询毕业证真伪的，而且国内没有渠道去查询国外文凭的真假，也不需要提供真实教育部认证。鉴于此，办理一份毕业证【微信：A575476】即可三、进国企，银行，事业单位，考公务员等等，这些单位是必需要提供真实教育部认证的，办理教育部认证所需资料众多且烦琐，所有材料您都必须提供原件，我们凭借丰富的经验，快捷的绿色通道帮您快速整合材料，让您少走弯路。留信网认证的作用: 1:该专业认证可证明留学生真实身份【微信：A575476】 2:同时对留学生所学专业登记给予评定 3:国家专业人才认证中心颁发入库证书 4:这个认证书并且可以归档倒地方 5:凡事获得留信网入网的信息将会逐步更新到个人身份内，将在公安局网内查询个人身份证信息后，同步读取人才网入库信息 6:个人职称评审加20分 7:个人信誉贷款加10分 8:在国家人才网主办的国家网络招聘大会中纳入资料，供国家高端企业选择人才 → 【关于价格问题（保证一手价格）我们所定的价格是非常合理的，而且我们现在做得单子大多数都是代理和回头客户介绍的所以一般现在有新的单子我给客户的都是第一手的代理价格，因为我想坦诚对待大家不想跟大家在价格方面浪费时间对于老客户或者被老客户介绍过来的朋友，我们都会适当给一些优惠。选择实体注册公司办理，更放心，更安全！我们的承诺：可来公司面谈，可签订合同，会陪同客户一起到教育部认证窗口递交认证材料，客户在教育部官方认证查询网站查询到认证通过结果后付款，不成功不收费！办理(USC毕业证）阳光海岸大学毕业证【微信：A575476】外观非常精致，由特殊纸质材料制成，上面印有校徽、校名、毕业生姓名、专业等信息。办理(USC毕业证）阳光海岸大学毕业证【微信：A575476】格式相对统一，各专业都有相应的模板。通常包括以下部分：校徽：象征着学校的荣誉和传承。校名:学校英文全称授予学位：本部分将注明获得的具体学位名称。毕业生姓名：这是最重要的信息之一，标志着该证书是由特定人员获得的。颁发日期：这是毕业正式生效的时间，也代表着毕业生学业的结束。其他信息：根据不同的专业和学位，可能会有一些特定的信息或章节。办理(USC毕业证）阳光海岸大学毕业证【微信：A575476】价值很高，需��妥善保管。一般来说，应放置在安全、干燥、防潮的地方，避免长时间暴露在阳光下。如需使用，最好使用复印件而不是原件，以免丢失。综上所述，办理(USC毕业证）阳光海岸大学毕业证【微信：A575476 】是证明身份和学历的高价值文件。外观简单庄重，格式统一，包括重要的个人信息和发布日期。对持有人来说，妥善保管是非常重要的。

Supervised Learning (Data Science).pptx

TARIKU ENDALE

Recently uploaded (20)

Streamlining Legacy Complexity Through Modernization

Malviya Nagar @ℂall @Girls ꧁❤ 9873940964 ❤꧂VIP Jina Singh Top Model Safe

BIGPPTTTTTTTTtttttttttttttttttttttt.pptx

[D3T2S03] Data&AI Roadshow 2024 - Amazon DocumentDB 실습

[D3T1S04] Aurora PostgreSQL performance monitoring and troubleshooting by use...

NPS_Presentation_V3.pptx it is regarding National pension scheme

Australian Catholic University degree offer diploma Transcript

EGU2020-10385_presentation LSTM algorithm

MUMBAI MONTHLY RAINFALL CAPSTONE PROJECT

University of Toronto degree offer diploma Transcript

From Clues to Connections: How Social Media Investigators Expose Hidden Networks

Rohini @ℂall @Girls ꧁❤ 9873940964 ❤꧂VIP Megha Singla Top Model Safe

LLM powered Contract Compliance Application.pptx

Sunshine Coast University diploma

Seamlessly Pay Online, Pay In Stores or Send Money

Noida Extension @ℂall @Girls ꧁❤ 9873940964 ❤꧂VIP Vishakha Singla Top Model Safe

iot paper presentation FINAL EDIT by kiran.pptx

Niagara College degree offer diploma Transcript

University of the Sunshine Coast degree offer diploma Transcript

Supervised Learning (Data Science).pptx

Building Real Time Targeting Capabilities - Ryan Zotti, Subbu Thiruppathy - Capital One

1. Building Real Time Targeting Capabilities Capital One | Fast Marketing July 20, 2016 | H20 Open Tour | NYC

2. Ryan Zotti Senior Data Engineer Subbu Thiruppathy Senior Software Engineer EXPERTISE FUN FACT Big Data, Python, R, Java, Machine Learning, AWS FUN FACT Big Data, Java, AKKA Play, AWS Built a self driving remote controlled car Recipient of Capital One’s most prestigious honor EXPERTISE

5. http://www.tibco.com/blog/2015/05/26/upcoming-webinar-integration-as-the-foundation-of-fast-data-may-28-2/

6. Source: http://thumbs.dreamstime.com/x/65-miles-per-hour-7772157.jpg https://morganalyx.wordpress.com/2013/02/22/assertive-driving/

7. http://www.dailymail.co.uk/health/article-2467478/What-causes-dry-eye-syndrome-cure-treatment.html

8. MODEL DATA MODEL DEPLOYMENT MODEL SCORING MODEL TRAINING Our challenge is… …striving to be fast at everything

9. Most current, up-to-date data Available as soon as it’s ready Low latency at scale FAST MODEL DATA

10. Most current, up-to-date data Available as soon as it’s ready Low latency at scale FAST MODEL DATA

11. Most current, up-to-date data Available as soon as it’s ready Low latency at scale FAST MODEL DATA

12. Most current, up-to-date data Available as soon as it’s ready Low latency at scale FAST MODEL DATA

13. Distributed computing to crunch data fast Elastic scaling with the public cloud Speed from parallelism FAST MODEL TRAINING

14. Distributed computing to crunch data fast Elastic scaling with the public cloud Speed from parallelism FAST MODEL TRAINING

15. Distributed computing to crunch data fast Elastic scaling with the public cloud Speed from parallelism FAST MODEL TRAINING

16. Distributed computing to crunch data fast Elastic scaling with the public cloud Speed from parallelism FAST MODEL TRAINING

17. Model adapts to evolving customer landscape Automatically refit the model and daily deploy Seamlessly integrate with existing Java tech stack FAST MODEL DEPLOYMENT

18. Model adapts to evolving customer landscape Automatically refit the model and daily deploy Seamlessly integrate with existing Java tech stack FAST MODEL DEPLOYMENT

19. Model adapts to evolving customer landscape Automatically refit the model and daily deploy Seamlessly integrate with existing Java tech stack FAST MODEL DEPLOYMENT

20. Model adapts to evolving customer landscape Automatically refit the model and daily deploy Seamlessly integrate with existing Java tech stack FAST MODEL DEPLOYMENT

21. Response < 100 milliseconds JVM-based model (i.e. POJO) Predictive power vs. runtime complexity (speed) Gradient boosting provided the best balance FAST MODEL SCORING

22. Response < 100 milliseconds JVM-based model (i.e. POJO) Predictive power vs. runtime complexity (speed) Gradient boosting provided the best balance FAST MODEL SCORING

23. Response < 100 milliseconds JVM-based model (i.e. POJO) Predictive power vs. runtime complexity (speed) Gradient boosting provided the best balance FAST MODEL SCORING

24. Response < 100 milliseconds JVM-based model (i.e. POJO) Predictive power vs. runtime complexity (speed) Gradient boosting provided the best balance FAST MODEL SCORING

25. Response < 100 milliseconds JVM-based model (i.e. POJO) Predictive power vs. runtime complexity (speed) Gradient boosting provided the best balance FAST MODEL SCORING

26. VISITOR WEBSITE API MODEL DATA

27. Explore new technologies continuously Ability to switch new models “on-the-fly” Make the API faster Incorporate new data sources Resiliency, failover capabilities

28. Technology changes Flexibility of the cloud Keep it simple Small empowered teams

29. THANK YOU

Editor's Notes

BOTH
Websites have evolved over time. And so has our website, check out what it looks like today! SUBBU
Capital One’s Credit Card Homepage has changed over time, so have our data and modeling needs SUBBU How can we make the card home page experience better Today everyone sees the same experience Future experience needs to be fast and targeted
How fast is fast data? RYAN
Tesla travels 0-60 MPH in 3.2 seconds RYAN
A blink of an eye takes 300-400 milliseconds RYAN
RYAN
RYAN
RYAN
RYAN
RYAN
RYAN
RYAN
RYAN
Solution: Sparking Water RYAN
RYAN
RYAN
RYAN
RYAN
RYAN
RYAN
RYAN
RYAN
RYAN
SUBBU
WHERE ARE WE GOING NEXT? SUBBU
Lessons Learned SUBBU Simple: network hops? Tech changes: Cloud: consumers are on cloud (faster vs on-prem), developer empowerment, on demand servers Small teams:

Building Real Time Targeting Capabilities - Ryan Zotti, Subbu Thiruppathy - Capital One

More Related Content

What's hot

What's hot (20)

Viewers also liked

Viewers also liked (20)

Similar to Building Real Time Targeting Capabilities - Ryan Zotti, Subbu Thiruppathy - Capital One

Similar to Building Real Time Targeting Capabilities - Ryan Zotti, Subbu Thiruppathy - Capital One (20)

More from Sri Ambati

More from Sri Ambati (20)

Recently uploaded

Recently uploaded (20)

Building Real Time Targeting Capabilities - Ryan Zotti, Subbu Thiruppathy - Capital One

Editor's Notes