Dive intro Deep Learning Book

The “Dive into Deep Learning Book” delves into the intricacies of deep learning, offering a comprehensive exploration of the subject matter. Through a series of detailed chapters, the book covers fundamental concepts, practical applications, and advanced techniques in the field of deep learning. Readers can expect to gain a solid understanding of neural networks, optimization algorithms, convolutional neural networks, recurrent neural networks, and more. With a focus on hands-on learning and real-world examples, this book serves as a valuable resource for both beginners and experienced practitioners looking to enhance their knowledge and skills in deep learning.

Visit Dive intro Deep Learning Book

Queries, Keys, and Values

So far all the networks we reviewed crucially relied on the input being of a well-defined size. For instance, the images in ImageNet are of size \(224 \times 224\) pixels and CNNs are specifically tun...

Reinforcement Learning

Dive intro Deep Learning Book

Pratik Chaudhari ( University of Pennsylvania and Amazon ), Rasool Fakoor ( Amazon ), and Kavosh Asadi ( Amazon ) Reinforcement Learning (RL) is a suite of techniques that allows us to build machine l...

Gaussian Process Inference

Dive intro Deep Learning Book

In this section, we will show how to perform posterior inference and make predictions using the GP priors we introduced in the last section. We will start with regression, where we can perform inferen...

Value Iteration

Dive intro Deep Learning Book

In this section we will discuss how to pick the best action for the robot at each state to maximize the return of the trajectory. We will describe an algorithm called Value Iteration and implement it ...

Q-Learning

Dive intro Deep Learning Book

In the previous section, we discussed the Value Iteration algorithm which requires accessing the complete Markov decision process (MDP), e.g., the transition and reward functions. In this section, we ...

Gaussian Processes

Dive intro Deep Learning Book

Andrew Gordon Wilson ( New York University and Amazon ) Gaussian processes (GPs) are ubitiquous. You have already encountered many examples of GPs without realizing it. Any model that is linear in its...

Introduction to Gaussian Processes

Dive intro Deep Learning Book

In many cases, machine learning amounts to estimating parameters from data. These parameters are often numerous and relatively uninterpretable — such as the weights of a neural network. Gaussian proce...

Gaussian Process Priors

Dive intro Deep Learning Book

Understanding Gaussian processes (GPs) is important for reasoning about model construction and generalization, and for achieving state-of-the-art performance in a variety of applications, including ac...

Markov Decision Process (MDP)

Dive intro Deep Learning Book

In this section, we will discuss how to formulate reinforcement learning problems using Markov decision processes (MDPs) and describe various components of MDPs in detail. 17.1.1. Definition of an MDP...

Hyperparameter Optimization API

Dive intro Deep Learning Book

Before we dive into the methodology, we will first discuss a basic code structure that allows us to efficiently implement various HPO algorithms. In general, all HPO algorithms considered here need to...

Asynchronous Random Search

Dive intro Deep Learning Book

As we have seen in the previous Section 19.2 , we might have to wait hours or even days before random search returns a good hyperparameter configuration, because of the expensive evaluation of hyperpa...

Hyperparameter Optimization

Dive intro Deep Learning Book

Aaron Klein ( Amazon ), Matthias Seeger ( Amazon ), and Cedric Archambeau ( Amazon ) The performance of every machine learning model depends on its hyperparameters. They control the learning algorithm...