AMLD2021 - ML in online marketplaces

•

0 likes•87 views

This document discusses the use of machine learning in online marketplaces. It outlines how machine learning is used for recommendations, search, trust and safety, seller experience, and pricing/monetization. Specific applications mentioned include collaborative and content-based recommendation systems, ranking models for search, automated content moderation, image quality assessment, dynamic pricing, and promoting listings. The document provides examples of algorithms like counting, collaborative filtering, learning to rank, and neural networks that power these machine learning applications in online marketplaces.

Machine Learning in
Online Marketplaces
Alexey Grigorev
14.09.2021

DataTalks.Club — @Al_Grigor
Plan
● Marketplace
● Examples of marketplaces
● Use cases of ML in marketplaces
○ Search
○ Recommendation
○ Trust and Safety
○ Seller experience
○ Pricing and monetization
○ Other examples
● Conclusion

DataTalks.Club — @Al_Grigor
Source: https://www.dvnovosti.ru/eao/2019/08/24/103683/

DataTalks.Club — @Al_Grigor
Seller
Icons: Bananas, Apple

DataTalks.Club — @Al_Grigor
Buyer Seller
Icons: Bananas, Apple

DataTalks.Club — @Al_Grigor
Buyer Seller
Icons: Bananas, Apple
Marketplace

DataTalks.Club — @Al_Grigor
Buyer Seller
Online
Marketplace
Icons: Bananas, Apple
Website
💳

DataTalks.Club — @Al_Grigor
Buyer
(Demand side)
Seller
(Supply side)
Two-sided
Marketplace
Website

DataTalks.Club — @Al_Grigor
ML in online marketplaces
● Recommendations
● Search
● Trust and safety
● Seller experience
● Pricing and monetization
● Other examples

DataTalks.Club — @Al_Grigor
Amazing 5G phone
With 100500 GB RAM
10 cameras
Unlimited disk space
Customers who bought this also bought ...
Icons: srip

DataTalks.Club — @Al_Grigor
Algorithms
● Counting
● Collaborative filtering
● Neural networks

DataTalks.Club — @Al_Grigor
Counting
SELECT country, item, count(1)
FROM views
GROUP BY country, item

DataTalks.Club — @Al_Grigor
Collaborative filtering
A
Users
Items
≈ U V
x

DataTalks.Club — @Al_Grigor
Collaborative filtering
Users
Items
≈ x
Item
vector
User
vector

DataTalks.Club — @Al_Grigor
Collaborative filtering
Icons: srip
similarity

DataTalks.Club — @Al_Grigor
Neural networks (“item2vec”)
Icons: srip
Neural network
Neural network
similarity

DataTalks.Club — @Al_Grigor
https://habr.com/ru/company/avito/blog/491942/
Deep Learning Recommender Systems - Cristian Martinez, Ilia Ivanov (video)

DataTalks.Club — @Al_Grigor
Cool devices
Icons: srip
🔍

DataTalks.Club — @Al_Grigor
Algorithms
● Bag of words and TF-IDF
● Learning to rank

DataTalks.Club — @Al_Grigor
Amazing 5G phone
With 100500 GB RAM
10 cameras
Unlimited disk space
Icons: srip
Apache
Lucene
Index

DataTalks.Club — @Al_Grigor
Amazing 5G phone
With 100500 GB RAM
10 cameras
Unlimited disk space
Icons: srip
Apache
Lucene
Cool devices
Index
Query

$DataTalks.Club — @Al_Grigor TF-IDF Q: {"cool": 1.1, "device": 1.2} D1: {"apple": 1.2, "5g": 1.0, "hip": 1.5} D2: {"samsung": 1.1, "joystick": 1.7, ...} D3: {"hp": 2.2, "laptop": 1.2, "used": 0.6, ...} D4: {"deep": 0.9, "learning": 1.0, "printer": 2.3, ...}$

DataTalks.Club — @Al_Grigor
doc
query
θ
angle between vectors
w2
w1

DataTalks.Club — @Al_Grigor
Learning to rank
Query 1 Document 1 Yes
Query 1 Document 2 No
Query 1 Document 3 No
Query 2 Document 1 No
Query 2 Document 2 No
Query 2 Document 3 Yes

DataTalks.Club �� @Al_Grigor
Query (user)
Document
Model P(relevant)

DataTalks.Club — @Al_Grigor
Query
Document
Neural network
Neural network
similarity

DataTalks.Club — @Al_Grigor
User generated content
Such description. So much text

DataTalks.Club — @Al_Grigor
User generated content
Such description. So
much text
Such description. So
much text
Such description. So
much text

DataTalks.Club — @Al_Grigor
User generated content

DataTalks.Club — @Al_Grigor
Problems:
● Illegal content
● NSFW content
● Duplicates
● Spam
● Fraud

DataTalks.Club — @Al_Grigor
Content moderation
Such description
So much text

DataTalks.Club — @Al_Grigor
Content moderation
ML
Such description
So much text
Automatic
moderation system

DataTalks.Club — @Al_Grigor
Content moderation
ML
Such description
So much text
Accept
Reject
Moderation queue
MP
Automatic
moderation system
Moderation panel
Accept
Reject
Moderators

DataTalks.Club — @Al_Grigor
ML
Automatic
moderation system
Duplicates
Forbidden
items
NSFW
Spam Fraud

DataTalks.Club — @Al_Grigor
category:
Truck
category:
Fridge
category:
Machete
Forbidden items

DataTalks.Club — @Al_Grigor
https://tech.olx.com/a-two-step-framework-for-duplicate-detection-fbbe4c905480
https://tech.olx.com/detecting-image-duplicates-at-olx-scale-7f59e4b6aef4

DataTalks.Club — @Al_Grigor
I want to
sell my
car
Picture by John Torcasio from unsplash (source)

DataTalks.Club — @Al_Grigor
Picture by John Torcasio from unsplash (source)
��

DataTalks.Club — @Al_Grigor
🚗 Model Price

DataTalks.Club — @Al_Grigor
quality:
Good
quality:
Bad
Image quality

DataTalks.Club — @Al_Grigor
AI in online marketplaces
● Recommendations
● Search
● Trust and safety
● Seller experience
● Pricing and monetization
● Other examples

DataTalks.Club — @Al_Grigor
��
🚗
Icons: Concert

DataTalks.Club — @Al_Grigor
��
🚗
Icons: Concert
��
��
��
��
🚗
🚗
“Surge”

DataTalks.Club — @Al_Grigor
Icons: srip
Cool devices
🚀
Awesome
Promoted
Ad
🔍

DataTalks.Club — @Al_Grigor
Plan
● Marketplace
● Examples of marketplaces
● Use cases of ML in marketplace
○ Search
○ Recommendation
○ Trust and Safety
○ Seller experience
○ Pricing and monetization
○ Other examples
● Conclusion

DataTalks.Club — @Al_Grigor
Conclusions
● ML in marketplaces is everywhere
● ML helps buyers: search, recommendation, trust and safety
● ML helps sellers: seller experience
● ML helps make money: pricing and monetization

DataTalks.Club — @Al_Grigor
@Al_Grigor
agrigorev
DataTalks.Club

Similar to AMLD2021 - ML in online marketplaces

Qcon SF 2013 - Machine Learning & Recommender Systems @ Netflix Scale

Xavier Amatriain

The document summarizes Netflix's approach to machine learning and recommender systems. It discusses how Netflix uses algorithms like SVD and Restricted Boltzmann Machines on a massive scale to power highly personalized recommendations. Over 75% of what people watch on Netflix comes from recommendations. Netflix collects a huge amount of data from over 40 million subscribers and uses both offline, online, and nearline computation across cloud services to train models and power recommendations in real-time at scale. The key is combining more data, smarter models, accurate metrics, and optimized system architectures.

Algorithm Marketplace and the new "Algorithm Economy"

Diego Oppenheimer

Diego Oppenheimer discusses the rise of algorithm marketplaces and the new "algorithm economy". Key points include: - Advances in machine learning, computer vision, speech recognition and natural language processing are enabling algorithms to interpret unstructured data at scale. - Algorithm marketplaces allow algorithms to be hosted, discovered, monetized and composed modularly to address a wide range of use cases across many industries. - The algorithm economy will lower barriers to applying machine intelligence and foster innovation as algorithms become reusable assets that creators and users can both benefit from.

Vlad Micu “20 things succesful game developers do beyond making games”

Lviv Startup Club

The document provides a 20-point list of things that successful game developers do beyond making games. It begins with an introduction to the author and his background. The main list includes recommendations such as prototyping quickly, ensuring the first 15 minutes of a game are engaging, preparing for business aspects early, gathering industry research, embracing analytics, and getting out of the trenches to network. Throughout are examples and the author's own experiences to illustrate each point.

How to Uncover Big Growth Opportunities with Data

Looker

Twitter, Airbnb, HotelTonight and Facebook all used insights found deep within their data to unlock big growth opportunities. Join us as Colin reverse engineers five data-driven growth wins and shows you how you can apply this same level of thinking and analysis to find growth for your company. In this webinar, Colin Zima, Chief Analytics Officer for Looker and Sean Ellis, CEO of Qualaroo, will show you how to find and tap into the hidden growth levers in your own data. Walk away with actionable strategies to transform your data into a growth engine for your business.

Data Science and Machine Learning for eCommerce and Retail

Andrei Lopatenko

This document discusses how machine learning and data science can be applied to ecommerce and retail. It outlines several machine learning problems that are relevant, such as search ranking, typeahead, spell correction, cold start recommendations, and inventory management. It also discusses the large data volumes and processing needs, as well as hardware requirements, for deploying machine learning solutions at large retailers.

AWS Activate Webinar - Growing on AWS

Amazon Web Services

This document discusses how a gaming startup can leverage analytics on Amazon Web Services to gain insights from log data. It provides examples of metrics that may be relevant at different stages of a startup, such as user engagement and revenue. It also outlines sources of log data on AWS, like CloudFront logs and web server logs, that can be used to measure metrics related to data transfer, top URLs, error rates, and more. The document emphasizes that startups should focus on one key metric tailored to their stage of growth and use log data to track and improve their product before funds run out.

Using Spark-Solr at Scale: Productionizing Spark for Search with Apache Solr...

Databricks

This talk is a case-study on how Apache Spark and the Spark-Solr library is being used at Flipp for driving search relevancy. Flipp is a Toronto based digital flyer and ecommerce company which helps shoppers save money on weekly shopping. Our customers have the option of browsing through our 5+ million products from the brick-and-mortar retailers in North America. This makes Search a very challenging function in our app. How to show the most relevant and personalized search results to users on a query? The talk will focus on using user signals such as Click Through Rate (CTR) and Impressions to increase search relevancy. I will also talk about how PySpark is used to create the Flipp Search ETL platform for collecting user signals and reading product data from Solr. The problem scenario will be explained in which keyword search and basic relevancy algorithms become ineffective when dealing with a large product database. The solutions will cover the following implementations being used at Flipp to drive relevancy: – Utilizing user clicks and popularity data to derive and index normalized item weights to implement the Search Crowd Curation models in Apache Solr – How around 5+ million items are classified into Google Categories in real time using Keras and Apache Spark to power product category curation in Solr. – How to create a crowd sourced query intent categorizer in Solr using the Spark-Solr library. – The use of offline and online metrics at Flipp for evaluating changes in search relevancy. – Future plans for incorporating Kafka-connect in Apache Solr with structured streaming to perform real-time product indexing with Spark-Solr library.

MLOps at OLX

Alexey Grigorev

This document discusses MLOps at OLX, including: - The main areas of data science work at OLX like search, recommendations, fraud detection, and content moderation. - How OLX uses teams structured by both feature areas and roles to collaborate on projects. - A maturity model for MLOps with levels from no MLOps to fully automated processes. - How OLX has improved from siloed work to cross-functional teams and adding more automation to model creation, release, and application integration over time.

NPTs

Brandon Levene

How to PM Hardware Products by Western Digital Sr. PM

Product School

ML Zoomcamp 1.4 - CRISP-DM

Alexey Grigorev

The document discusses the CRISP-DM methodology for machine learning projects. It describes the six main steps: 1) business understanding to define goals, 2) data understanding to analyze sources, 3) data preparation to transform data, 4) modeling to select the best model, 5) evaluation to validate goals are met, and 6) deployment to production. These steps provide an iterative process from problem definition to model deployment.

Machine Learning at Netflix Scale

Aish Fenton

Netflix is the world’s leading Internet television network with over 48 million members in more than 40 countries enjoying more than one billion hours of TV shows and movies per month, including original series. Netflix uses machine learning to deliver a personalized experience to each one of our 48 million users. In this talk you will hear about the machine learning algorithms that power almost every part of the Netflix experience, including some of our recent work on distributed Neural Networks on AWS GPUs. You will also get an insight into the innovation approach that includes offline experimentation and online AB testing. Finally, you will learn about the system architectures that enable all of this at a Netflix scale.

A whirlwind tour of graph databases

jexp

Value Driven Regulatory Use Case Simulation

Aaron Benningfield

This presentation is designed to provide concepts, architecture and use cases for capturing hidden value and new opportunities by integrating current & new regulations along with disparate sources of data (e.g., legal proceedings, other regulatory requirements, future industry considerations) into a company’s IT systems. A framework is provided for organizations to develop a business and technical strategy to generate a host of benefits such as revenue streams, new opportunities, knowledge bases and intellectual property. Please feel free to contact me for an depth discussion.

SystemT: Declarative Information Extraction (invited talk at MIT CSAIL)

Laura Chiticariu

Invited talk at MIT CSAIL, March 8 2016 Information extraction (IE), the task of extracting structured information from unstructured or semi-structured data, is increasingly important to a wide array of enterprise applications, ranging from Business Intelligence to Data-as-a-Service. Such applications drive the following main requirements for IE systems: accuracy, scalability, expressivity, transparency, and customizability. SystemT, a declarative IE system, has been designed and developed to address these requirements. It is based on the basic principle underlying relational database technology: complete separation of specification from execution. SystemT uses a declarative language for expressing NLP algorithms called AQL, and an optimizer that generates high-performance algebraic execution plans for AQL rules. It makes IE orders of magnitude more scalable and easy to use, maintain and customize. SystemT ships today with multiple products across 4 IBM Software Brands. Furthermore, SystemT is used in multiple ongoing research projects and being taught in universities. Our ongoing research and development efforts focus on making SystemT more usable for both technical and business users, and continuing enhancing its core functionalities based on natural language processing, machine learning, and database technology.

20 Things Successful Game Developers Do Beyond Making Games

Vlad Micu

Real-Time Fraud Detection at Scale—Integrating Real-Time Deep-Link Graph Anal...

Databricks

This document discusses using TigerGraph for real-time fraud detection at scale by integrating real-time deep-link graph analytics with Spark AI. It provides examples of common TigerGraph use cases including recommendation engines, fraud detection, and risk assessment. It then discusses how TigerGraph can power explainable AI by extracting over 100 graph-based features from entities and their relationships to feed machine learning models. Finally, it shares a case study of how China Mobile used TigerGraph for real-time phone-based fraud detection by analyzing over 600 million phone numbers and 15 billion call connections as a graph to detect various types of fraud in real-time.

Auto ai for skillsfuture

Sunny Panjabi

This document introduces H2O.ai and their product H2O Driverless AI, an automated machine learning platform. It provides an overview of the company and product, including use cases across different industries where customers saw improvements in key metrics like accuracy, time savings, and performance. It also shares resources for learning more about Driverless AI through tutorials, documentation, and community forums.

Crafting Expertise, Authority and Trust with Entity-Based Content Strategy - ...

Jamie Indigo

At SMXL, I presented a talk about crafting effective, authoritative content by understanding entities. People, places, objects, and ideas have facets. Human users have unique perspectives and their language changes as their relationship to an entity changes. It's time we stop chasing keywords-- a byproduct of search intent-- in favor of strategic entity-based strategy. This deck includes insights into how to access the data behind Google's knowledge graph, use external links to boost the search engine's understanding, and ways to become an authoritative and trusted source.

Data, AI, and Tokens: Ocean Protocol

Trent McConaghy

This talk introduces Ocean protocol. It describes: -how data drives AI (artificial intelligence) -the gap between data-haves and AI-haves -the data silo crisis -how Ocean addresses these issues by creating a substrate to catalyze a flowering of data marketplaces -the Ocean structured approach to token design, from values to stakeholders to software stack. Video: https://www.youtube.com/watch?v=fMDD0aTVt4s This talk was presented at "9984 Summit - Blockchain Futures for Developers, Enterprises, and Society" hosted by IPDB & BigchainDB.

Similar to AMLD2021 - ML in online marketplaces (20)

Qcon SF 2013 - Machine Learning & Recommender Systems @ Netflix Scale

Algorithm Marketplace and the new "Algorithm Economy"

Vlad Micu “20 things succesful game developers do beyond making games”

How to Uncover Big Growth Opportunities with Data

Data Science and Machine Learning for eCommerce and Retail

AWS Activate Webinar - Growing on AWS

Using Spark-Solr at Scale: Productionizing Spark for Search with Apache Solr...

MLOps at OLX

NPTs

How to PM Hardware Products by Western Digital Sr. PM

ML Zoomcamp 1.4 - CRISP-DM

Machine Learning at Netflix Scale

A whirlwind tour of graph databases

Value Driven Regulatory Use Case Simulation

SystemT: Declarative Information Extraction (invited talk at MIT CSAIL)

20 Things Successful Game Developers Do Beyond Making Games

Real-Time Fraud Detection at Scale—Integrating Real-Time Deep-Link Graph Anal...

Auto ai for skillsfuture

Crafting Expertise, Authority and Trust with Entity-Based Content Strategy - ...

Data, AI, and Tokens: Ocean Protocol

More from Alexey Grigorev

MLOps week 1 intro

Alexey Grigorev

Data Monitoring with whylogs

Alexey Grigorev

Whylogs is an open source tool for data monitoring that automatically creates statistical summaries called profiles of datasets. It helps with data monitoring by generating these profiles which can be compared over time to detect changes visually or programmatically. This allows issues like schema changes or bugs in data pipelines to be identified. The profiles have properties like being descriptive, lightweight and mergeable, which enables monitoring across distributed systems by allowing profile data to be logically merged. Whylogs thus provides a step towards observability of data systems.

Data engineering zoomcamp introduction

Alexey Grigorev

The document outlines the plan and syllabus for a Data Engineering Zoomcamp hosted by DataTalks.Club. It introduces the four instructors for the course - Ankush Khanna, Sejal Vaidya, Victoria Perez Mola, and Alexey Grigorev. The 10-week course will cover topics like data ingestion, data warehousing with BigQuery, analytics engineering with dbt, batch processing with Spark, streaming with Kafka, and a culminating 3-week student project. Pre-requisites include experience with Python, SQL, and the command line. Course materials will be pre-recorded videos and there will be weekly live office hours for support. Students can earn a certificate and compete on a

AI in Fashion - Size & Fit - Nour Karessli

Alexey Grigorev

This document discusses Zalando's use of AI to improve size and fit recommendations for customers. It outlines several challenges including varying size conventions, limited fit data for new items, and sparse customer purchase histories. It then describes Zalando's approaches to address these, including algorithms that use item images to predict sizes for new items lacking data (SizeNet) and models that learn from customers' past purchases and feedback to provide personalized size recommendations. The goal is to help customers find the right fit on their first purchase to reduce returns and improve the shopping experience.

AI-Powered Computer Vision Applications in Media Industry - Yulia Pavlova

Alexey Grigorev

Computer vision techniques like facial recognition and image captioning can help automate metadata generation for media companies. Facial recognition can identify people in photos to assist editors and improve searchability, while image captioning can propose captions. A case study of applying these techniques to photos from English Premier League football games achieved 99% accuracy for facial recognition and precision of 78.7% for image captioning. Combining the two allows generating customized captions that include names identified through facial recognition. Challenges remain when the automatic caption does not match details in the image.

ML Zoomcamp 10 - Kubernetes

Alexey Grigorev

Paradoxes in Data Science

Alexey Grigorev

This document discusses several paradoxes that can arise in data science. It begins by discussing modelling and simulations that can be used when data is unavailable. It then outlines Simpson's Paradox, where a trend seen in groups disappears or reverses when the groups are combined. Next, it discusses the accuracy paradox, where a metric stops being useful once it becomes the target. It also discusses the learnability-Godel paradox related to the limitations of mathematics according to Godel's incompleteness theorems. Finally, it discusses the law of unintended consequences as it relates to data science.

ML Zoomcamp 8 - Neural networks and deep learning

Alexey Grigorev

Algorithmic fairness

Alexey Grigorev

An algorithm is considered fair if its results and performance are independent of sensitive variables like gender, ethnicity, etc. Fairness can be introduced at different stages of model development, such as in data collection, preparation, and model selection. Techniques for identifying and mitigating bias include causal reasoning, explainability, fairness metrics, and counterfactuals. Counterfactual fairness evaluates predictions across different protected attribute values while holding other variables constant. Explainability helps ensure models make decisions for the right reasons. Overall fairness aims to achieve equal outcomes or opportunities across groups.

ML Zoomcamp 6 - Decision Trees and Ensemble Learning

Alexey Grigorev

ML Zoomcamp 5 - Model deployment

Alexey Grigorev

Introduction to Transformers for NLP - Olga Petrova

Alexey Grigorev

Olga Petrova gives an introduction to transformers for natural language processing (NLP). She begins with an overview of representing words using tokenization, word embeddings, and one-hot encodings. Recurrent neural networks (RNNs) are discussed as they are important for modeling sequential data like text, but they struggle with long-term dependencies. Attention mechanisms were developed to address this by allowing the model to focus on relevant parts of the input. Transformers use self-attention and have achieved state-of-the-art results in many NLP tasks. Bidirectional Encoder Representations from Transformers (BERT) provides contextualized word embeddings trained on large corpora.

ML Zoomcamp 4 - Evaluation Metrics for Classification

Alexey Grigorev

ML Zoomcamp 3 - Machine Learning for Classification

Alexey Grigorev

ML Zoomcamp Week #2 Office Hours

Alexey Grigorev

ML Zoomcamp 2 - Slides

Alexey Grigorev

ML Zoomcamp - Course Overview and Logistics

Alexey Grigorev

This document outlines the course plan for a Machine Learning zoom camp hosted by DataTalks.Club. The 11-session course will cover topics ranging from introductory machine learning concepts to advanced techniques like deep learning, model deployment, and Kubernetes. Participants should have some Python experience and be comfortable with command line. Completing homework assignments and projects can earn participants up to 100 points to be listed on a public leaderboard. The goal is to help attendees learn applied machine learning skills in a collaborative, public setting.

ML Zoomcamp 1.10 - Summary

Alexey Grigorev

This document contains summaries from multiple sessions of a machine learning zoomcamp. It introduces machine learning concepts like supervised learning, the CRISP-DM process, model selection, linear algebra, and the Python libraries NumPy and Pandas. It also discusses setting up an environment for machine learning and provides example data and models for tasks like email spam detection and car price prediction.

ML Zoomcamp 1.8 - Linear Algebra Refresher

Alexey Grigorev

ML Zoomcamp 1.5 - Model Selection Process

Alexey Grigorev

The document discusses the process of selecting the best machine learning model, including evaluating multiple models on a holdout dataset to determine the best performing one. It notes common model types like logistic regression, decision trees, and neural networks that could be considered. It also addresses the need to split data into training, validation, and test sets to accurately assess model performance and avoid overfitting conclusions from multiple comparisons of models during selection.

More from Alexey Grigorev (20)

MLOps week 1 intro

Data Monitoring with whylogs

Data engineering zoomcamp introduction

AI in Fashion - Size & Fit - Nour Karessli

AI-Powered Computer Vision Applications in Media Industry - Yulia Pavlova

ML Zoomcamp 10 - Kubernetes

Paradoxes in Data Science

ML Zoomcamp 8 - Neural networks and deep learning

Algorithmic fairness

ML Zoomcamp 6 - Decision Trees and Ensemble Learning

ML Zoomcamp 5 - Model deployment

Introduction to Transformers for NLP - Olga Petrova

ML Zoomcamp 4 - Evaluation Metrics for Classification

ML Zoomcamp 3 - Machine Learning for Classification

ML Zoomcamp Week #2 Office Hours

ML Zoomcamp 2 - Slides

ML Zoomcamp - Course Overview and Logistics

ML Zoomcamp 1.10 - Summary

ML Zoomcamp 1.8 - Linear Algebra Refresher

ML Zoomcamp 1.5 - Model Selection Process

Recently uploaded

Exploring Deep Learning Models for Image Recognition: A Comparative Review

sipij

Image recognition, which comes under Artificial Intelligence (AI) is a critical aspect of computer vision, enabling computers or other computing devices to identify and categorize objects within images. Among numerous fields of life, food processing is an important area, in which image processing plays a vital role, both for producers and consumers. This study focuses on the binary classification of strawberries, where images are sorted into one of two categories. We Utilized a dataset of strawberry images for this study; we aim to determine the effectiveness of different models in identifying whether an image contains strawberries. This research has practical applications in fields such as agriculture and quality control. We compared various popular deep learning models, including MobileNetV2, Convolutional Neural Networks (CNN), and DenseNet121, for binary classification of strawberry images. The accuracy achieved by MobileNetV2 is 96.7%, CNN is 99.8%, and DenseNet121 is 93.6%. Through rigorous testing and analysis, our results demonstrate that CNN outperforms the other models in this task. In the future, the deep learning models can be evaluated on a richer and larger number of images (datasets) for better/improved results.

UNIT I INCEPTION OF INFORMATION DESIGN 20CDE09-ID

GOWSIKRAJA PALANISAMY

Rotary Intersection in traffic engineering.pptx

surekha1287

Development of Chatbot Using AI/ML Technologies

maisnampibarel

The rapid advancements in artificial intelligence and natural language processing have significantly transformed human-computer interactions. This thesis presents the design, development, and evaluation of an intelligent chatbot capable of engaging in natural and meaningful conversations with users. The chatbot leverages state-of-the-art deep learning techniques, including transformer-based architectures, to understand and generate human-like responses. Key contributions of this research include the implementation of a context- aware conversational model that can maintain coherent dialogue over extended interactions. The chatbot's performance is evaluated through both automated metrics and user studies, demonstrating its effectiveness in various applications such as customer service, mental health support, and educational assistance. Additionally, ethical considerations and potential biases in chatbot responses are examined to ensure the responsible deployment of this technology. The findings of this thesis highlight the potential of intelligent chatbots to enhance user experience and provide valuable insights for future developments in conversational AI.

Software Engineering and Project Management - Introduction to Project Management

Prakhyath Rai

Introduction to Project Management: Introduction, Project and Importance of Project Management, Contract Management, Activities Covered by Software Project Management, Plans, Methods and Methodologies, some ways of categorizing Software Projects, Stakeholders, Setting Objectives, Business Case, Project Success and Failure, Management and Management Control, Project Management life cycle, Traditional versus Modern Project Management Practices.

L-3536-Cost Benifit Analysis in ESIA.pptx

naseki5964

Chlorine and Nitric Acid application, properties, impacts.pptx

yadavsuyash008

Best Practices of Clothing Businesses in Talavera, Nueva Ecija, A Foundation ...

IJAEMSJORNAL

This study primarily aimed to determine the best practices of clothing businesses to use it as a foundation of strategic business advancements. Moreover, the frequency with which the business's best practices are tracked, which best practices are the most targeted of the apparel firms to be retained, and how does best practices can be used as strategic business advancement. The respondents of the study is the owners of clothing businesses in Talavera, Nueva Ecija. Data were collected and analyzed using a quantitative approach and utilizing a descriptive research design. Unveiling best practices of clothing businesses as a foundation for strategic business advancement through statistical analysis: frequency and percentage, and weighted means analyzing the data in terms of identifying the most to the least important performance indicators of the businesses among all of the variables. Based on the survey conducted on clothing businesses in Talavera, Nueva Ecija, several best practices emerge across different areas of business operations. These practices are categorized into three main sections, section one being the Business Profile and Legal Requirements, followed by the tracking of indicators in terms of Product, Place, Promotion, and Price, and Key Performance Indicators (KPIs) covering finance, marketing, production, technical, and distribution aspects. The research study delved into identifying the core best practices of clothing businesses, serving as a strategic guide for their advancement. Through meticulous analysis, several key findings emerged. Firstly, prioritizing product factors, such as maintaining optimal stock levels and maximizing customer satisfaction, was deemed essential for driving sales and fostering loyalty. Additionally, selecting the right store location was crucial for visibility and accessibility, directly impacting footfall and sales. Vigilance towards competitors and demographic shifts was highlighted as essential for maintaining relevance. Understanding the relationship between marketing spend and customer acquisition proved pivotal for optimizing budgets and achieving a higher ROI. Strategic analysis of profit margins across clothing items emerged as crucial for maximizing profitability and revenue. Creating a positive customer experience, investing in employee training, and implementing effective inventory management practices were also identified as critical success factors. In essence, these findings underscored the holistic approach needed for sustainable growth in the clothing business, emphasizing the importance of product management, marketing strategies, customer experience, and operational efficiency.

21CV61- Module 3 (CONSTRUCTION MANAGEMENT AND ENTREPRENEURSHIP.pptx

sanabts249

Introduction to IP address concept - Computer Networking

Md.Shohel Rana ( M.Sc in CSE Khulna University of Engineering & Technology (KUET))

An Internet Protocol address (IP address) is a logical numeric address that is assigned to every single computer, printer, switch, router, tablets, smartphones or any other device that is part of a TCP/IP-based network. Types of IP address- Dynamic means "constantly changing “ .dynamic IP addresses aren't more powerful, but they can change. Static means staying the same. Static. Stand. Stable. Yes, static IP addresses don't change. Most IP addresses assigned today by Internet Service Providers are dynamic IP addresses. It's more cost effective for the ISP and you.

Germany Offshore Wind 010724 RE (1) 2 test.pptx

rebecca841358

Advances in Detect and Avoid for Unmanned Aircraft Systems and Advanced Air M...

VICTOR MAESTRE RAMIREZ

kiln burning and kiln burner system for clinker

hamedmustafa094

Quadcopter Dynamics, Stability and Control

Blesson Easo Varghese

Response & Safe AI at Summer School of AI at IIITH

IIIT Hyderabad

Unit 1 Information Storage and Retrieval

KishorMahale5

Natural Is The Best: Model-Agnostic Code Simplification for Pre-trained Large...

YanKing2

Pre-trained Large Language Models (LLM) have achieved remarkable successes in several domains. However, code-oriented LLMs are often heavy in computational complexity, and quadratically with the length of the input code sequence. Toward simplifying the input program of an LLM, the state-of-the-art approach has the strategies to filter the input code tokens based on the attention scores given by the LLM. The decision to simplify the input program should not rely on the attention patterns of an LLM, as these patterns are influenced by both the model architecture and the pre-training dataset. Since the model and dataset are part of the solution domain, not the problem domain where the input program belongs, the outcome may differ when the model is trained on a different dataset. We propose SlimCode, a model-agnostic code simplification solution for LLMs that depends on the nature of input code tokens. As an empirical study on the LLMs including CodeBERT, CodeT5, and GPT-4 for two main tasks: code search and summarization. We reported that 1) the reduction ratio of code has a linear-like relation with the saving ratio on training time, 2) the impact of categorized tokens on code simplification can vary significantly, 3) the impact of categorized tokens on code simplification is task-specific but model-agnostic, and 4) the above findings hold for the paradigm–prompt engineering and interactive in-context learning and this study can save reduce the cost of invoking GPT-4 by 24%per API query. Importantly, SlimCode simplifies the input code with its greedy strategy and can obtain at most 133 times faster than the state-of-the-art technique with a significant improvement. This paper calls for a new direction on code-based, model-agnostic code simplification solutions to further empower LLMs.

Paharganj @ℂall @Girls ꧁❤ 9873777170 ❤꧂VIP Arti Singh Top Model Safe

aarusi sexy model

IWISS Catalog 2024

Iwiss Tools Co.,Ltd

Profiling of Cafe Business in Talavera, Nueva Ecija: A Basis for Development ...

IJAEMSJORNAL

This study aimed to profile the coffee shops in Talavera, Nueva Ecija, to develop a standardized checklist for aspiring entrepreneurs. The researchers surveyed 10 coffee shop owners in the municipality of Talavera. Through surveys, the researchers delved into the Owner's Demographic, Business details, Financial Requirements, and other requirements needed to consider starting up a coffee shop. Furthermore, through accurate analysis, the data obtained from the coffee shop owners are arranged to derive key insights. By analyzing this data, the study identifies best practices associated with start-up coffee shops’ profitability in Talavera. These findings were translated into a standardized checklist outlining essential procedures including the lists of equipment needed, financial requirements, and the Traditional and Social Media Marketing techniques. This standardized checklist served as a valuable tool for aspiring and existing coffee shop owners in Talavera, streamlining operations, ensuring consistency, and contributing to business success.

Recently uploaded (20)

Exploring Deep Learning Models for Image Recognition: A Comparative Review

UNIT I INCEPTION OF INFORMATION DESIGN 20CDE09-ID

Rotary Intersection in traffic engineering.pptx

Development of Chatbot Using AI/ML Technologies

Software Engineering and Project Management - Introduction to Project Management

L-3536-Cost Benifit Analysis in ESIA.pptx

Chlorine and Nitric Acid application, properties, impacts.pptx

Best Practices of Clothing Businesses in Talavera, Nueva Ecija, A Foundation ...

21CV61- Module 3 (CONSTRUCTION MANAGEMENT AND ENTREPRENEURSHIP.pptx

Introduction to IP address concept - Computer Networking

Germany Offshore Wind 010724 RE (1) 2 test.pptx

Advances in Detect and Avoid for Unmanned Aircraft Systems and Advanced Air M...

kiln burning and kiln burner system for clinker

Quadcopter Dynamics, Stability and Control

Response & Safe AI at Summer School of AI at IIITH

Unit 1 Information Storage and Retrieval

Natural Is The Best: Model-Agnostic Code Simplification for Pre-trained Large...

Paharganj @ℂall @Girls ꧁❤ 9873777170 ❤꧂VIP Arti Singh Top Model Safe

IWISS Catalog 2024

Profiling of Cafe Business in Talavera, Nueva Ecija: A Basis for Development ...

AMLD2021 - ML in online marketplaces

1. Machine Learning in Online Marketplaces Alexey Grigorev 14.09.2021

2. Hello 👋 I’m Alexey.

3. DataTalks.Club — @Al_Grigor Plan ● Marketplace ● Examples of marketplaces ● Use cases of ML in marketplaces ○ Search ○ Recommendation ○ Trust and Safety ○ Seller experience ○ Pricing and monetization ○ Other examples ● Conclusion

4. DataTalks.Club — @Al_Grigor Source: https://www.dvnovosti.ru/eao/2019/08/24/103683/

5. DataTalks.Club — @Al_Grigor Seller

6. DataTalks.Club — @Al_Grigor Seller Icons: Bananas, Apple

7. DataTalks.Club — @Al_Grigor Buyer Seller Icons: Bananas, Apple

8. DataTalks.Club — @Al_Grigor Buyer Seller Icons: Bananas, Apple 💶 💶 💶

9. DataTalks.Club — @Al_Grigor Buyer Seller Icons: Bananas, Apple 💶 💶 💶

10. DataTalks.Club — @Al_Grigor Buyer Seller Icons: Bananas, Apple Marketplace

11. DataTalks.Club — @Al_Grigor Buyer Seller Online Marketplace Icons: Bananas, Apple Website 💳

12. DataTalks.Club — @Al_Grigor Buyer (Demand side) Seller (Supply side) Two-sided Marketplace Website

14. DataTalks.Club — @Al_Grigor

15. DataTalks.Club — @Al_Grigor

16. DataTalks.Club — @Al_Grigor

17. DataTalks.Club — @Al_Grigor

18. DataTalks.Club — @Al_Grigor

19. DataTalks.Club — @Al_Grigor ML in online marketplaces ● Recommendations ● Search ● Trust and safety ● Seller experience ● Pricing and monetization ● Other examples

20. DataTalks.Club — @Al_Grigor

21. DataTalks.Club — @Al_Grigor Icons: srip

22. DataTalks.Club — @Al_Grigor Amazing 5G phone With 100500 GB RAM 10 cameras Unlimited disk space Customers who bought this also bought ... Icons: srip

23. DataTalks.Club — @Al_Grigor Algorithms ● Counting ● Collaborative filtering ● Neural networks

24. DataTalks.Club — @Al_Grigor Icons: srip

25. DataTalks.Club — @Al_Grigor Counting SELECT country, item, count(1) FROM views GROUP BY country, item

26. DataTalks.Club — @Al_Grigor Collaborative filtering A Users Items ≈ U V x

27. DataTalks.Club — @Al_Grigor Collaborative filtering Users Items ≈ x Item vector User vector

28. DataTalks.Club — @Al_Grigor d2 d1 󰬋 🧦 🧸

29. DataTalks.Club — @Al_Grigor d2 d1 󰬋 🧦 🧸

30. DataTalks.Club — @Al_Grigor Amazing 5G phone With 100500 GB RAM 10 cameras Unlimited disk space Customers who bought this also bought ... Icons: srip

31. DataTalks.Club — @Al_Grigor Collaborative filtering Icons: srip similarity

32. DataTalks.Club — @Al_Grigor Neural networks (“item2vec”) Icons: srip Neural network Neural network similarity

33. DataTalks.Club — @Al_Grigor https://habr.com/ru/company/avito/blog/491942/ Deep Learning Recommender Systems - Cristian Martinez, Ilia Ivanov (video)

34. DataTalks.Club — @Al_Grigor ML in online marketplaces ● Recommendations ● Search ● Trust and safety ● Seller experience ● Pricing and monetization ● Other examples

35. DataTalks.Club — @Al_Grigor Cool devices Icons: srip 🔍

36. DataTalks.Club — @Al_Grigor Algorithms ● Bag of words and TF-IDF ● Learning to rank

37. DataTalks.Club — @Al_Grigor Amazing 5G phone With 100500 GB RAM 10 cameras Unlimited disk space Icons: srip Apache Lucene Index

38. DataTalks.Club — @Al_Grigor Amazing 5G phone With 100500 GB RAM 10 cameras Unlimited disk space Icons: srip Apache Lucene Cool devices Index Query

39. DataTalks.Club — @Al_Grigor TF-IDF Q: {"cool": 1.1, "device": 1.2} D1: {"apple": 1.2, "5g": 1.0, "hip": 1.5} D2: {"samsung": 1.1, "joystick": 1.7, ...} D3: {"hp": 2.2, "laptop": 1.2, "used": 0.6, ...} D4: {"deep": 0.9, "learning": 1.0, "printer": 2.3, ...}

40. DataTalks.Club — @Al_Grigor doc query θ angle between vectors w2 w1

41. DataTalks.Club — @Al_Grigor Learning to rank Query 1 Document 1 Yes Query 1 Document 2 No Query 1 Document 3 No Query 2 Document 1 No Query 2 Document 2 No Query 2 Document 3 Yes

42. DataTalks.Club — @Al_Grigor Query (user) Document Model P(relevant)

43. DataTalks.Club — @Al_Grigor Query Document Neural network Neural network similarity

44. DataTalks.Club — @Al_Grigor ML in online marketplaces ● Recommendations ● Search ● Trust and safety ● Seller experience ● Pricing and monetization ● Other examples

45. DataTalks.Club — @Al_Grigor User generated content Such description. So much text

46. DataTalks.Club — @Al_Grigor User generated content Such description. So much text Such description. So much text Such description. So much text

47. DataTalks.Club — @Al_Grigor User generated content

48. DataTalks.Club — @Al_Grigor Problems: ● Illegal content ● NSFW content ● Duplicates ● Spam ● Fraud

49. DataTalks.Club — @Al_Grigor Content moderation Such description So much text

50. DataTalks.Club — @Al_Grigor Content moderation ML Such description So much text Automatic moderation system

51. DataTalks.Club — @Al_Grigor Content moderation ML Such description So much text Accept Reject Moderation queue MP Automatic moderation system Moderation panel Accept Reject Moderators

52. DataTalks.Club — @Al_Grigor ML Automatic moderation system Duplicates Forbidden items NSFW Spam Fraud

53. DataTalks.Club — @Al_Grigor category: Truck category: Fridge category: Machete Forbidden items

54. DataTalks.Club — @Al_Grigor https://tech.olx.com/a-two-step-framework-for-duplicate-detection-fbbe4c905480 https://tech.olx.com/detecting-image-duplicates-at-olx-scale-7f59e4b6aef4

55. DataTalks.Club — @Al_Grigor ML in online marketplaces ● Recommendations ● Search ● Trust and safety ● Seller experience ● Pricing and monetization ● Other examples

56. DataTalks.Club — @Al_Grigor I want to sell my car Picture by John Torcasio from unsplash (source)

57. DataTalks.Club — @Al_Grigor Picture by John Torcasio from unsplash (source) ��

58. DataTalks.Club — @Al_Grigor ��

59. DataTalks.Club — @Al_Grigor ��

60. DataTalks.Club — @Al_Grigor 🚗 Model Price

61. DataTalks.Club — @Al_Grigor �� 🥳 50000

62. DataTalks.Club — @Al_Grigor quality: Good quality: Bad Image quality

63. DataTalks.Club — @Al_Grigor AI in online marketplaces ● Recommendations ● Search ● Trust and safety ● Seller experience ● Pricing and monetization ● Other examples

64. DataTalks.Club — @Al_Grigor ��

65. DataTalks.Club — @Al_Grigor �� 🚗

66. DataTalks.Club — @Al_Grigor �� 🚗 Icons: Concert

67. DataTalks.Club — @Al_Grigor �� 🚗 Icons: Concert �� 🚗 🚗 “Surge”

68. DataTalks.Club — @Al_Grigor Icons: srip Cool devices 🚀 Awesome Promoted Ad 🔍

69. DataTalks.Club — @Al_Grigor Plan ● Marketplace ● Examples of marketplaces ● Use cases of ML in marketplace ○ Search ○ Recommendation ○ Trust and Safety ○ Seller experience ○ Pricing and monetization ○ Other examples ● Conclusion

70. DataTalks.Club — @Al_Grigor Conclusions ● ML in marketplaces is everywhere ● ML helps buyers: search, recommendation, trust and safety ● ML helps sellers: seller experience ● ML helps make money: pricing and monetization

71. DataTalks.Club — @Al_Grigor @Al_Grigor agrigorev DataTalks.Club

AMLD2021 - ML in online marketplaces

More Related Content

Similar to AMLD2021 - ML in online marketplaces

Similar to AMLD2021 - ML in online marketplaces (20)

More from Alexey Grigorev

More from Alexey Grigorev (20)

Recently uploaded

Recently uploaded (20)

AMLD2021 - ML in online marketplaces