Questions tagged [recommendation-engine]
For questions relating to recommendation engines, collaborative filtering, and personalization. Questions tend to be algorithmic or statistical in nature.
recommendation-engine
1,450
questions
0
votes
0
answers
15
views
Azure recommendation "Upgrade to the latest Azure CLI version" for VMSS at subscription level
I have been seeing a recommendation showing up across all of my account subscriptions suggesting "Upgrade to the latest Azure CLI version". I haven't got any VMs but on Linux VMSS under our ...
-1
votes
0
answers
10
views
Hybridized collaborative filtering and sentence similarity-based system for doctor recommendation based on user input of symptoms and location
I'm trying to solve a problem of recommending a doctor based on a user's symptoms and location using a hybridized collaborative filtering and sentence similarity-based recommender system that follow ...
0
votes
1
answer
28
views
ValueError: Wrong number of items passed 9, placement implies 1
I am trying to calculate the cosine similarity between two columns in a dataframe. the code snippet for it is as follows :
def cal_cosine_similarity(row):
vec1 = np.array(row['sup_vec'])
vec2 =...
0
votes
1
answer
108
views
Polars - issues with performance - attempting to create a new dataframe per row
I need to run another library's algo on each row of a large df but am having trouble converting my code to polars expressions for better performance. Here are a couple sample DFs:
df_products = pl....
0
votes
0
answers
26
views
Recommender System - Singular Value Decomposition (SVD) is providing random results
Sometimes, the quality of the output is simply assessed by eyeballing it. Looking at the examples provided below, if we use simple intuition it's clear that the expected user ratings are 2 and 3 (...
0
votes
1
answer
34
views
Memory Error with Cosine Similarity on Large Matrix (47605 x 73875) in Jupyter Notebook
I'm working on a recommendation system using python in Jupyter Notebook and need to compute cosine similarity for a large count matrix using sklearn.metrics.pairwise.cosine_similarity. However, I'm ...
0
votes
0
answers
61
views
UserWarning: LightFM was compiled without OpenMP support
Please help! I'm trying to make a movie recommendation system.
I'm on windows using PyCharm.
I've tried installing lightfm on both PIP and Conda.
Any suggestions are welcome! I'm only a beginner.
Code:...
0
votes
0
answers
72
views
Adding user and item features to LightFM Model
I am training a LightFM model for a recommender system, I create my interactions dataset successfully then try to add user and item features but face a consistent error stating the following (...
-1
votes
1
answer
45
views
Recommendations in Elasticsearch
There is a product database; the popularity value for products is periodically calculated. An index has been built in ES, there is sorting by popularity value.
There is a task to screw up a ...
0
votes
0
answers
16
views
Pytorch Embedding Layers RecSys
class RecommendationSystemModel(nn.Module):
def __init__(
self,
num_users,
num_movies,
embedding_size=128,
hidden_dim=128,
dropout_rate=0.2,
):
...
0
votes
0
answers
8
views
In the training of KGCN, how are the project feature vectors and user feature vectors updated, and how is the loss of the model updated?
I'm having trouble reading the KGCN open source code used for recommendations.In the train method of the source code, the following statements were used for training the model:
# KGCN is a custom ...
0
votes
0
answers
34
views
Pytorch got Non-reproducible results after adding activation function (i.e., ReLU)
I am unable to reproduce my results in PyTorch after adding a nn.ReLU(). I am sure that the problem must be here instead of other places, since I have tested with ablation hundreds of times. It is ...
0
votes
0
answers
9
views
How do I leverage Shopify APIs to do catalogue pull and order push of the brands that I have partnered with?
I want to create an SDK that allows any kind of a platform (content / media / OTT) to be able to specify what kind of brands they want to show and allow them to showcase it on their website. For this, ...
0
votes
0
answers
10
views
Using different hierarchical features for training and prediction
Im creating a recommender system that takes in customer data(ids and 5 features) and product data(ids and 1 feature). It predicts the revenue for each product from a given customer. The product ...
0
votes
0
answers
22
views
Data Cardinality is Ambiguous when running model.fit for a matrix factorization model
I'm making a recommender system using a matrix factorization model. Here is how I define the model:
import tensorflow as tf
from keras.layers import Add, Dense, Dot, Embedding, Reshape, Input
from ...