Description

This comprehensive course will help you learn how to use the power of Python to evaluate your deep learning-based recommender system data sets based on user ratings and choices with a practical approach to building a deep learning-based recommender system by adopting a retrieval-based approach based on a two-tower model.

Recommender systems are used in various areas with commonly recognized examples, including playlist generators for video and music services, product recommenders for online stores and social media platforms, and open web content recommenders. Recommender systems have also been developed to explore research articles and experts, collaborators, and financial services. The course begins with an introduction to deep learning concepts to develop recommender systems and a course overview. The course advances to topics covered, including deep learning for recommender systems, understanding the pros and cons of deep learning, recommendation inference, and deep learning-based recommendation approach. You will then explore neural collaborative filtering and learn how to build a project based on the Amazon Product Recommendation System. You will learn to install the required packages, analyze data for products recommendation, prepare data, and model development using a two-tower approach. You will learn to implement a TensorFlow recommender and test a recommender model. You will make predictions using the built recommender system. Upon completion, you can relate the concepts and theories for recommender systems in various domains and implement deep learning models for building real-world recommendation systems. All resources are available at: https://github.com/PacktPublishing/Recommender-Systems-An-Applied-Approach-using-Deep-Learning

What You Will Learn

Learn about deep learning and recommender systems
Explore the mechanisms of deep learning-based approaches
Learn to implement a two-tower model for recommenders
Implement TensorFlow to develop a recommender system
Learn basic neural network models for recommendations
Explore neural collaborative filtering and variational autoencoders

Audience

This course is designed for individuals looking to advance their skills in applied deep learning, understand relationships of data analysis with deep learning, build customized recommender systems for their applications, and implement deep learning algorithms for recommender systems. Individuals passionate about recommender systems with the help of TensorFlow Recommenders will benefit from this course. Deep learning practitioners, research scholars, and data scientists will also benefit from the course. The prerequisites include a basic to intermediate knowledge of Python and Pandas library.

Approach

The course is well-structured and explanatory. Every module has engaging content covering necessary theoretical concepts with practical explanations. The lectures are divided into many videos and comprehensively detailed code notebooks, providing a unique hands-on experience using a real-time project. This course is easily understandable, expressive, and self-explanatory, with live coding.

Key Features

Understand, implement, and evaluate deep learning models for building real-world recommendation systems * Validate, test, and make predictions using recommender systems with the help of TensorFlow Recommenders * Explore the benefits and challenges of deep learning in recommender systems

Github Repo

https://github.com/PacktPublishing/Recommender-Systems-An-Applied-Approach-using-Deep-Learning

Course Outline

1. Introduction

This section focuses on introducing the author and presenting a generalized overview of the course content.

1. About the Instructor

In this video, we will get a brief outline of the author, AI Sciences, and how they deliver their courses.

2. Course Outline

This brief introductory video will cover the course contents and the concepts you will learn in this course.

2. Deep Learning Foundation for Recommender Systems

This section focuses on deep learning methodologies used in recommender systems. Here, we will cover inferences, embeddings, and neutral collaborative filtering aside from VAE and deep learning models.

1. Module Introduction

In this video, we will look at a general introduction to deep learning and what is covered in this section under deep learning.

2. Overview

This video briefly overviews the contents and topics covered in this module, including deep learning in recommender systems, the pros and cons of deep learning, recommender interference, and more.

3. Deep Learning in Recommender Systems

In this video, we will explore deep learning concepts in recommender systems, migrating from machine learning to deep learning, and capturing non-linear to non-trivial relationships.

4. Inference after Training

In this lecture, you will learn how to deploy the model as a service to infer the likelihood of new interactions.

5. Inference Mechanism

In this lesson, we will understand the generic recommendation systems interference mechanism.

6. Embeddings and User Context

This video focuses on collaborative filtering with the help of deep learning and neural collaborative filtering.

7. Neural Collaborative Filtering

In this video, we will understand neural collaborative filtering and how to incorporate it into our recommender system.

8. VAE Collaborative Filtering

This lesson explores the variational autoencoder for collaborative filtering, using representations in hidden layers.

9. Strengths and Weaknesses of DL Models

In this lesson, you will learn about the strengths and weaknesses of deep learning recommender system models.

10. Deep Learning Quiz

This quiz is based on the deep learning concepts learned so far in this module.

11. Deep Learning Quiz Solution

This is the solution to the quiz based on the deep learning concepts learned so far in this module.

3. Project Amazon Product Recommendation System

This module focuses on TensorFlow recommenders, creating a two-tower model, data visualization with WordCloud, creating tensors with DataFrame, query towers, random train test split, and computer loss.

1. Module Overview

This is a brief introductory overview of what is being covered in this section, including creating projects, TensorFlow recommenders, libraries, WordCloud, and DataFrame.

2. TensorFlow Recommenders

In this video, you will learn about TensorFlow recommenders used to develop recommender systems.

3. Two-Tower Model

In this lesson, we will discuss the two-tower model, which uses user embedding and item embedding in the recommender system.

4. Project Overview

This video outlines the project and will demonstrate how to install the packages required for the project, data preparation, using TensorFlow, evaluation, and recommendation.

5. Download Libraries

In this video, you will learn to implement the Amazon recommender system using the TensorFlow recommender based on specific user behavior. You will learn to download and import particular libraries for the project.

6. Data Visualization with WordCloud

In this video, you will learn how to load a dataset for the project being developed, using WordCloud, importing WordCloud, STOPWORDS, and ImageColorGenerator.

7. Make Tensors from DataFrame

In this video, we will continue to check our dataset using a single username. We will use a part of the DataFrame to do so.

8. Rating Our Data

In this video, you will learn about the next part of the project, which is rating our data.

9. Random Train-Test Split

In this lesson, you will learn to perform the test train split, which will do the training split and then create a prediction; we will use a random dataset (80-20 ratio).

10. Making the Model and Query Tower

In this lecture, you will learn to develop our model and create a query tower to perform retrieval tasks.

11. Candidate Tower and Retrieval System

In this video, you will learn how to create a candidate tower and develop a retrieval system.

12. Compute Loss

Here, we will look at the next step of training our model: the compute loss function.

13. Train and Validation

After entirely developing our model, you will learn to fit and evaluate the model for functionality.

14. Accuracy Versus Recommendations

In this video, we will perform data visualization with the project we created and check the model's accuracy.

15. Making Recommendations

In this lesson, you will learn to create a recommendation and use a brute-force algorithm to generate the recommendation.

Course Images

Recommender Systems: An Applied Approach using Deep Learning

By Packt

Booking options

Highlights

Description

Course Content

About The Provider

Tags

Reviews