Building Recommender Systems with Machine Learning and AI

1. Installing Anaconda, Course Materials, and Creating Movie Recommendations

This video demonstrates how to install Anaconda, highlights the course materials, and explains how to create movie recommendations.

2. Course Roadmap

This video provides the structure of the course and explains how to navigate through the course.

3. What Is a Recommender System?

This video introduces you to the concept of a recommender system.

4. Types of Recommenders

This video focuses on the type of recommenders.

5. Understanding the Implicit and Explicit Ratings

This video explains how the recommender system learns about individual tastes and preferences. It also explains how explicit and implicit ratings work, their strengths, and their weaknesses.

6. Top-N Recommender Architecture

This video introduces you to the top-N recommender architecture and explains a few architectural approaches to build top-N systems.

7. Review the Basics of a Recommender System

This video provides a review of the basic recommender system with a quick four-question quiz along with the answers.

1. The Basics of Python (Activity)

This video explains the basics of Python.

2. Data Structures in Python

This video introduces you to data structures in Python, such as lists, tuples, and dictionaries.

3. Functions in Python

This video demonstrates how to define a function in Python.

4. Booleans, Loops, and a Hands-on Challenge (Exercise)

This video explains how Boolean expressions and loops work in Python. It also provides a challenge to write a simple Python function.

1. Train/Test and Cross-Validation

This video explains how to perform train/test and cross-validation in a recommender system.

2. Accuracy Metrics Using Root Mean Squared Error (RMSE) and Mean Absolute Error (MAE)

This video explains how to find accuracy using the RMSE and MAE measures and highlights the need to use these measures of recommendation prediction accuracy.

3. Top-N Hit Rate - Many Ways

This video focuses on several ways to measure the accuracy of top-N recommenders, including hit rate, cumulative hit rate, average reciprocal hit rank, rating hit rate, and many more.

4. Coverage, Diversity, and Novelty

This video demonstrates how to measure the coverage of your recommender system and highlights how diverse and novel its results are.

5. Churn, Responsiveness, and A/B Tests

This video explains how to measure your recommendations change (churn,) how quickly they respond to new data (responsiveness), and why no metric matters more than the results of real, online A/B tests.

6. Reviewing Ways to Measure Your Recommender (Quiz)

This video provides a short quiz to review the different ways to measure the qualities and accuracy of your recommender system.

7. Walkthrough of RecommenderMetrics.py (Activity)

This video provides a walkthrough of this course's Python module for implementing the metrics on the recommender system.

8. Walkthrough of TestMetrics.py (Activity)

This video provides a walkthrough of TestMetrics.py.

9. Measuring the Performance of Singular Value Decomposition (SVD) Recommendations (Activity)

This video demonstrates how to measure the performance of a SVD recommender and explains how to interpret them.

1. Recommender Engine Architecture

This video focuses on the architecture of the recommender engine framework, which helps to easily implement, test, and compare different algorithms throughout this course.

2. Recommender Engine Walkthrough - Part 1 (Activity)

This is the first part of the two-part video that provides a code walkthrough of the recommender engine.

3. Recommender Engine Walkthrough - Part 2 (Activity)

This is the second part of the two-part video that provides a code walkthrough of the recommender engine.

4. Reviewing the Results of Our Algorithm Evaluation (Activity)

This video demonstrates how to review the results of the framework to evaluate the Singular Value Decomposition (SVD) algorithm and interpret them.

1. Content-Based Recommendations and the Cosine Similarity Metric

This video focuses on content-based recommendations and the cosine similarity metric.

2. K-Nearest-Neighbors (KNN) and Content Recommendations

This video explains the concept of KNN and content recommendations.

3. Producing and Evaluating Content-Based Movie Recommendations (Activity)

This video shows how to produce and evaluate content-based movie recommendations based on their genres and years and evaluate the results using the MovieLens dataset.

4. A Note on Using Implicit Ratings

This video explains how to use implicit ratings, such as a purchase or click data.

5. Bleeding Edge Alert! Mise-En-Scene Recommendations (Activity)

This video explains how to use the mise-en-scene data for providing additional content-based information to our recommendations.

6. Diving Deep into Content-Based Recommendations (Exercise)

This video explains the content-based recommendations in detail and how to augment content-based recommendations using popularity data.

1. Measuring Similarity and Sparsity

This video explains how to measure similarity and sparsity.

2. Similarity Metrics

This video shows the different ways of measuring similarity, including cosine, adjusted cosine, Pearson, Spearman, Jaccard, and many more.

3. User-Based Collaborative Filtering

This video highlights the concept of user-based collaborative filtering.

4. User-Based Collaborative Filtering - Hands-On (Activity)

This video demonstrates how to write some code to apply user-based collaborative filtering to the MovieLens dataset.

5. Item-Based Collaborative Filtering

This video explains the item-based collaborative filtering and how it works.

6. Item-Based Collaborative Filtering - Hands-On (Activity)

This video shows how to write, run, and evaluate code to apply item-based collaborative filtering to generate recommendations from the MovieLens dataset and compare it to the user-based collaborative filtering.

7. Tuning Collaborative Filtering Algorithms (Exercise)

This video demonstrates how to improve the user-based and item-based collaborative filtering algorithms by tweaking the way candidate generation works.

8. Evaluating Collaborative Filtering Systems Offline (Activity)

This video explains how to evaluate collaborative filtering systems offline.

9. Measuring the Hit Rate of Item-based Collaborative Filtering (Exercise)

This video shows how to measure the hit rate of item-based collaborative filtering.

10. K-Nearest-Neighbor (KNN) Recommenders

This video demonstrates how to apply the ideas of neighborhood-based collaborative filtering into frameworks based on rating predictions with KNN recommenders.

11. Running User-Based and Item-Based K-Nearest-Neighbor (KNN) on MovieLens Dataset (Activity)

This video shows how to run a user-based and item-based KNN on MovieLens dataset and evaluate the results.

12. Experimenting with Different K-Nearest-Neighbor (KNN) Parameters (Exercise)

This is an exercise video focuses on the different K-Nearest-Neighbor (KNN) parameters.

13. Bleeding Edge Alert! Translation-Based Recommendations

This video provides a "bleeding edge alert" to discuss the translation-based recommendations.

1. Principal Component Analysis (PCA)

This video introduces you to PCA.

2. Singular Value Decomposition (SVD)

This video introduces you to SVD.

3. Running Singular Value Decomposition (SVD) and Singular Value Decomposition (SVD ++) on MovieLens (Activity)

This video demonstrates how to run SVD and SVD++ on MovieLens movie rating dataset and evaluate the results.

4. Improving on Singular Value Decomposition (SVD)

This video focuses on improving SVD.

5. Tuning the Hyperparameters on Singular Value Decomposition (SVD) (Exercise)

This video demonstrates how to modify the SVD bake-off code to find the optimal values of the various hyperparameters for SVD and see if it makes a difference in the results.

6. Bleeding-Edge Alert! Sparse Linear Methods (SLIM)

This video provides a bleeding edge alert on SLIM.

1. Deep Learning Introduction

This video provides an introduction to deep learning.

2. Deep Learning Prerequisites

This video focuses on the prerequisites required for deep learning.

3. History of Artificial Neural Networks (ANN)

This video explains the history of ANN.

4. (Activity) Playing with TensorFlow

This video demonstrates how to use the TensorFlow playground to get a hands-on feel of how deep neural networks operate and the effects of different topologies.

5. Training Neural Networks

This video focuses on the mechanism of different activation functions and optimization functions for neural networks, including ReLU, Adam, RMSProp, and gradient descent.

6. Tuning Neural Networks (Avoiding Overfitting)

This video focuses on how to prevent overfitting using techniques such as dropout layers, and how to tune your topology for the best results.

7. Activation Functions: More Depth

This video explains the activation functions in detail.

8. Introduction to TensorFlow

This video introduces you to TensorFlow.

9. Handwriting Recognition with TensorFlow - Part 1 (Activity)

This is the first part of the two-part video that demonstrates how to implement handwriting recognition on real data using TensorFlow's low-level API.

10. Handwriting Recognition with TensorFlow - Part 2 (Activity)

This is the second part of the two-part video that explains how to implement handwriting recognition on real data using TensorFlow's low-level API.

11. Introduction to Keras

This video introduces you to Keras, a higher-level API that makes developing deep neural networks with TensorFlow a lot easier.

12. Handwriting Recognition with Keras (Activity)

This video demonstrates how to tackle the handwriting recognition using Keras API with much simpler code and better results.

13. Classifier Patterns with Keras

This video explains the different patterns to use in Keras API for multi-class or binary classification problems.

14. Predicting Political Parties of Politicians with Keras (Exercise)

This video provides a challenge to develop a neural network using Keras API to predict the political parties of politicians.

15. Introduction to Convolutional Neural Network (CNN)

This video provides an introduction to CNN.

16. Convolutional Neural Network (CNN) Architecture

This video focuses on the architecture of CNN.

17. Handwriting Recognition with Convolutional Neural Network (CNN) (Activity)

This video explains how to tackle handwriting recognition using CNN.

18. Introduction to Recurrent Neural Networks (RNN)

This video introduces you to RNN. They are appropriate for sequences of information, such as time-series data, natural language, or music.

19. Training Recurrent Neural Networks (RNN)

This video demonstrates how to train recurrent RNN that involves back-propagating through time.

20. Sentiment Analysis of Movie Reviews Using Recurrent Neural Networks (RNN) and Keras (Activity)

This video shows how to apply RNN to the problem of sentiment analysis, which can be modeled as a sequence-to-vector learning problem.

21. Tuning Neural Networks

This video explains how to tune neural networks.

22. Neural Network Regularization Techniques

This video focuses on neural network regularization techniques.

23. Generative Adversarial Networks (GANs)

This video explains about the Generative Adversarial Networks (GANs).

24. GANs in Action

This video demonstrates GANs in action.

25. Generating Images of Clothing with Generative Adversarial Networks (Activity)

This video demonstrates how to generate images of clothing using the Generative Adversarial Networks (GANs).

1. Introduction to Deep Learning for Recommenders

This video provides an introduction to deep learning for recommenders.

2. Restricted Boltzmann Machine (RBM)

This video explains RBM using the recommender system.

3. Recommendations with Restricted Boltzmann Machine (RBM) - Part 1 (Activity)

This is the first part of the two-part video that explains how to build recommendations using RBM.

4. Recommendations with Restricted Boltzmann Machine (RBM) - Part 2 (Activity)

This is the second part of the two-part video that explains how to build recommendations using RBM.

5. Evaluating the Restricted Boltzmann Machine (RBM) Recommender (Activity)

This video focuses on evaluating the RBM recommender and studying its results.

6. Tuning Restricted Boltzmann Machine (RBM) (Exercise)

This video provides a challenge to tune the RBM using GridSearchCV.

7. Exercise Results: Tuning a Restricted Boltzmann Machine (RBM) Recommender

This video provides the results from the previous exercise to compare with the challenge results.

8. Auto-Encoders for Recommendations: Deep Learning for Recommendation

This video explains how to apply modern deep neural networks to the recommender system using auto-encoder.

9. Recommendations with Deep Neural Networks (Activity)

This video explains how to build recommendations using deep neural networks.

10. Clickstream Recommendations with Recurrent Neural Networks (RNN)

This video introduces you to "GRU4Rec," a technique that applies RNN to the problem of clickstream recommendations.

11. Getting GRU4Rec to Work on Your Desktop (Exercise)

This video provides a challenge to get GRU4Rec to work on the desktop.

12. Exercise Results: GRU4Rec in Action

This video reviews the results from the previous exercise.

13. Bleeding Edge Alert! Generative Adversarial Networks for Recommendations

This video explains the concept of Generative Adversarial Networks for Recommendations.

14. TensorFlow Recommenders (TFRS): Introduction and Building a Retrieval Stage

This video talks about the TensorFlow Recommenders (TFRS): Introduction and Building a Retrieval Stage.

15. TensorFlow Recommenders (TFRS): Building a Ranking Stage

This video talks about the TensorFlow Recommenders (TFRS): Building a Ranking Stage.

16. TensorFlow Recommenders (TFRS): Incorporating Side Features and Deep Retrieval

This video talks about the TensorFlow Recommenders (TFRS): Incorporating Side Features and Deep Retrieval.

17. TensorFlow Recommenders (TFRS): Multi-Task Recommenders, Deep and Cross Networks, ScaNN, and Serving

This video talks about the TensorFlow Recommenders (TFRS): Multi-Task Recommenders, Deep and Cross Networks, ScaNN, and Serving.

18. Bleeding Edge Alert! Deep Factorization Machines

This video explains the concept of deep factorization machine.

19. More Emerging Tech to Watch

This video focuses on a few more "bleeding edge" topics, including Word2Vec, 3D CNN's for session-based recommendations and feature extraction with CNNs.

1. Introduction and Installation of Apache Spark (Activity)

This video introduces you to Apache Spark and explains how to install it on the system.

2. Apache Spark Architecture

This video focuses on the Apache Spark architecture.

3. Movie Recommendations with Spark, Matrix Factorization, and Alternating Least Squares (ALS) (Activity)

This video demonstrates how to use Spark's MLLib to generate recommendations with ALS.

4. Recommendations from 20 Million Ratings with Spark (Activity)

This video explains how to process 20 million ratings and produce top-N recommendations with Apache Spark.

5. Amazon Deep Scalable Sparse Tensor Network Engine (DSSTNE)

This video explains the recommender engine called DSSTNE, which makes it easy to apply deep neural networks to massive, sparse datasets and produce great recommendations at a large scale.

6. Amazon Deep Scalable Sparse Tensor Network Engine (DSSTNE) in Action

This video demonstrates how to use Amazon DSSTNE on an EC2 Ubuntu instance to produce movie recommendations using a deep neural network.

7. Scaling Up Amazon Deep Scalable Sparse Tensor Network Engine (DSSTNE)

This video shows how Amazon scaled up DSSTNE and paired with Apache Spark to process their massive data and produce recommendations for millions of customers.

8. Amazon Web Services (AWS) SageMaker and Factorization Machines

This video introduces you to Amazon's SageMaker service that offers machine learning algorithms that can be used for recommendations, including factorization machines.

9. Amazon SageMaker in Action: Factorization Machines on One Million Ratings in the Cloud

This video demonstrates how to use the Amazon SageMaker from a cloud-hosted notebook to pre-process the MovieLens one-million-rating dataset.

10. Other Systems of Note (Amazon Personalize, RichRelevance, Recombee, and More)

This video introduces you to other system of note, such as Amazon Personalize, RichRelevance, Recombee, and many more.

11. Recommender System Architecture

This video explains the recommender system architecture that specifies how to deploy a recommender system into production based on the environment.

1. The Cold Start Problem (and Solutions)

This video demonstrates how to make recommendations for a brand-new user or with brand-new items.

2. Implementing Random Exploration (Exercise)

This video explains how to implement random exploration of new items using underutilized slots in recommendation results.

3. Exercise Solution - Random Exploration

This video provides a walkthrough of random exploration with the results.

4. Stoplists

This video explains the concept of stoplists.

5. Implementing a Stoplist (Exercise)

This video provides a challenge on how to implement a stoplist to keep movies with potentially offensive titles out of your results entirely.

6. Exercise Solution - Implementing a Stoplist

This video provides a solution for implementing a stoplist.

7. Filtering Bubbles, Trust, and Outliers

This video shows how to filter bubbles, trust, and outliers.

8. Identifying and Eliminating Outlier Users (Exercise)

This video explains how to identify and eliminate outlier users.

9. Exercise Solution: Outlier Removal

This video provides a review of the implementation of outlier detection and removal along with the results.

10. Fraud, the Perils of Clickstream, and International Concerns

This video demonstrates how to prevent fraud, explains the perils of clickstream, and focuses on international concerns recommendations.

11. Temporal Effects and Value-Aware Recommendations

This video presents an overview of temporal effects and value-aware recommendations.

1. Case Study: YouTube - Part 1

This video explains the challenges unique to recommendations at YouTube and highlights their high-level approach using deep learning.

2. Case Study: YouTube - Part 2

This video provides a case study to dive deeper into YouTube architecture for applying deep learning to both candidate generation and candidate ranking.

3. Case Study: Netflix - Part 1

This video focuses on the Netflix recommendations, their approach, and heavy use of hybrid algorithms.

4. Case Study: Netflix - Part 2

This video provides a case study that explains how Netflix produces context-aware recommendations.

1. Hybrid Recommenders and Exercise

This video explains the concept of hybrid recommenders and provides a challenge to build a hybrid algorithm in the recommender framework that can combine any list of algorithms into one.

2. Exercise Solution: Hybrid Recommenders

This video focuses on the hybrid algorithm implementation and its results.

1. More to Explore

This video provides pointers to external books, conferences, and papers to keep you up-to-date after this course.