From the course: Building Recommender Systems with Machine Learning and AI
Unlock this course with a free trial
Join today to access over 23,200 courses taught by industry experts.
Case study: YouTube, part 2 - Python Tutorial
From the course: Building Recommender Systems with Machine Learning and AI
Case study: YouTube, part 2
- [Instructor] YouTube was kind enough to publish their deep learning architecture, at least as it was in 2016. Let's start at the bottom, where we have the user behavior data that is used to train the system. It's interesting that, although YouTube has explicit ratings in the form of thumbs-up thumbs-down ratings, they don't use them at all for generating recommendations because that data is too sparse. Not enough users rate videos explicitly for the data to be useful. Instead, they rely on implicit signals, such as which videos you actually watched and what you searched for. This implicit view in search data, however, is in itself sparse. And as we learned, when covering deep learning recommenders, dealing with that sparsity is a huge issue when trying to apply deep learning to recommender systems. Their solution was to break up the sparse representation of video IDs and search tokens for each user into a variable length sequence of sparse data, mapped to a dense layer of a fixed…
Practice while you learn with exercise files
Download the files the instructor uses to teach the course. Follow along and learn by watching, listening and practicing.
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.
Contents
-
-
(Locked)
Train/test and cross-validation3m 49s
-
(Locked)
Accuracy metrics (RMSE and MAE)4m 6s
-
Top-N hit rate: Many ways4m 35s
-
(Locked)
Coverage, diversity, and novelty4m 55s
-
(Locked)
Churn, responsiveness, and A/B tests5m 6s
-
(Locked)
Review ways to measure your recommender2m 55s
-
(Locked)
Walkthrough of RecommenderMetrics.py6m 53s
-
(Locked)
Walkthrough of TestMetrics.py5m 8s
-
(Locked)
Measure the performance of SVD recommendations2m 24s
-
(Locked)
-
-
(Locked)
Content-based recommendations and the cosine similarity metric8m 58s
-
K-nearest neighbors (KNN) and content recs3m 59s
-
(Locked)
Producing and evaluating content-based movie recommendations5m 23s
-
(Locked)
Bleeding edge alert: Mise-en-scene recommendations4m 31s
-
(Locked)
Dive deeper into content-based recommendations4m 26s
-
(Locked)
-
-
(Locked)
Measuring similarity and sparsity4m 49s
-
(Locked)
Similarity metrics8m 32s
-
User-based collaborative filtering7m 25s
-
(Locked)
User-based collaborative filtering: Hands-on4m 59s
-
(Locked)
Item-based collaborative filtering4m 14s
-
(Locked)
Item-based collaborative filtering: Hands-on2m 23s
-
(Locked)
Tuning collaborative filtering algorithms3m 31s
-
(Locked)
Evaluating collaborative filtering systems offline1m 28s
-
(Locked)
Measure the hit rate of item-based collaborative filtering2m 17s
-
(Locked)
KNN recommenders4m 4s
-
(Locked)
Running user- and item-based KNN on MovieLens2m 26s
-
(Locked)
Experiment with different KNN parameters4m 25s
-
(Locked)
Bleeding edge alert: Translation-based recommendations2m 29s
-
(Locked)
-
-
(Locked)
Deep learning introduction1m 30s
-
(Locked)
Deep learning prerequisites8m 13s
-
History of artificial neural networks10m 51s
-
(Locked)
Playing with TensorFlow12m 2s
-
(Locked)
Training neural networks5m 47s
-
(Locked)
Tuning neural networks3m 52s
-
(Locked)
Introduction to TensorFlow11m 29s
-
(Locked)
Handwriting recognition with TensorFlow, part 113m 18s
-
(Locked)
Handwriting recognition with TensorFlow, part 212m 3s
-
(Locked)
Introduction to Keras2m 48s
-
(Locked)
Handwriting recognition with Keras9m 52s
-
(Locked)
Classifier patterns with Keras3m 58s
-
(Locked)
Predict political parties of politicians with Keras9m 55s
-
(Locked)
Intro to convolutional neural networks (CNNs)8m 59s
-
(Locked)
CNN architectures2m 54s
-
(Locked)
Handwriting recognition with CNNs8m 38s
-
(Locked)
Intro to recurrent neural networks (RNNs)7m 38s
-
(Locked)
Training recurrent neural networks3m 21s
-
(Locked)
Sentiment analysis of movie reviews using RNNs and Keras11m 1s
-
(Locked)
-
-
(Locked)
Intro to deep learning for recommenders2m 19s
-
(Locked)
Restricted Boltzmann machines (RBMs)8m 2s
-
(Locked)
Recommendations with RBMs, part 112m 46s
-
Recommendations with RBMs, part 27m 11s
-
(Locked)
Evaluating the RBM recommender3m 44s
-
(Locked)
Tuning restricted Boltzmann machines1m 43s
-
(Locked)
Exercise results: Tuning a RBM recommender1m 15s
-
(Locked)
Auto-encoders for recommendations: Deep learning for recs4m 27s
-
(Locked)
Recommendations with deep neural networks7m 23s
-
(Locked)
Clickstream recommendations with RNNs7m 23s
-
(Locked)
Get GRU4Rec working on your desktop2m 42s
-
(Locked)
Exercise results: GRU4Rec in action7m 51s
-
(Locked)
Bleeding edge alert: Deep factorization machines5m 49s
-
(Locked)
More emerging tech to watch5m 14s
-
(Locked)
-
-
(Locked)
Introduction and installation of Apache Spark5m 49s
-
(Locked)
Apache Spark architecture5m 13s
-
(Locked)
Movie recommendations with Spark, matrix factorization, and ALS6m 2s
-
(Locked)
Recommendations from 20 million ratings with Spark4m 57s
-
(Locked)
Amazon DSSTNE4m 41s
-
DSSTNE in action9m 25s
-
(Locked)
Scaling up DSSTNE2m 14s
-
(Locked)
AWS SageMaker and factorization machines4m 24s
-
(Locked)
SageMaker in action: Factorization machines on one million ratings, in the cloud7m 39s
-
(Locked)
-
-
(Locked)
The cold start problem (and solutions)6m 12s
-
(Locked)
Implement random exploration54s
-
(Locked)
Exercise solution: Random exploration2m 18s
-
(Locked)
Stoplists4m 48s
-
(Locked)
Implement a stoplist32s
-
(Locked)
Exercise solution: Implement a stoplist2m 22s
-
(Locked)
Filter bubbles, trust, and outliers5m 39s
-
(Locked)
Identify and eliminate outlier users44s
-
(Locked)
Exercise solution: Outlier removal4m
-
Fraud, the perils of clickstream, and international concerns4m 33s
-
(Locked)
Temporal effects and value-aware recommendations3m 30s
-
(Locked)