book

Machine Learning with TensorFlow, Second Edition

Name: Machine Learning with TensorFlow, Second Edition
Author: Chris Mattmann
ISBN: 9781617297717

by Chris Mattmann

January 2021

Intermediate to advanced

456 pages

14h 7m

English

Manning Publications

Read now

Unlock full access

Machine Learning with TensorFlow, 2e
Copyright
dedication
Praise for the First Edition
front matter
forewordprefaceacknowledgmentsabout this bookHow this book is organized: A roadmapAbout the codeliveBook discussion forumabout the authorabout the cover illustration
contents
Part 1 Your machine-learning rig
1 A machine-learning odyssey
1.1 Machine-learning fundamentals1.1.1 Parameters1.1.2 Learning and inference1.2 Data representation and features1.3 Distance metrics1.4 Types of learning1.4.1 Supervised learning1.4.2 Unsupervised learning1.4.3 Reinforcement learning1.4.4 Meta-learning1.5 TensorFlow1.6 Overview of future chaptersSummary
2 TensorFlow essentials
2.1 Ensuring that TensorFlow works2.2 Representing tensors2.3 Creating operators2.4 Executing operators within sessions2.5 Understanding code as a graph2.5.1 Setting session configurations2.6 Writing code in Jupyter2.7 Using variables2.8 Saving and loading variables2.9 Visualizing data using TensorBoard2.9.1 Implementing a moving average2.9.2 Visualizing the moving average2.10 Putting it all together: The TensorFlow system architecture and APISummary
Part 2 Core learning algorithms

3 Linear regression and beyond
3.1 Formal notation3.1.1 How do you know the regression algorithm is working?3.2 Linear regression3.3 Polynomial model3.4 Regularization3.5 Application of linear regressionSummary
4 Using regression for call-center volume prediction
4.1 What is 311?4.2 Cleaning the data for regression4.3 What’s in a bell curve? Predicting Gaussian distributions4.4 Training your call prediction regressor4.5 Visualizing the results and plotting the error4.6 Regularization and training test splitsSummary
5 A gentle introduction to classification
5.1 Formal notation5.2 Measuring performance5.2.1 Accuracy5.2.2 Precision and recall5.2.3 Receiver operating characteristic curve5.3 Using linear regression for classification5.4 Using logistic regression5.4.1 Solving 1D logistic regression5.4.2 Solving 2D regression5.5 Multiclass classifier5.5.1 One-versus-all5.5.2 One-versus-one5.5.3 Softmax regression5.6 Application of classificationSummary
6 Sentiment classification: Large movie-review dataset
6.1 Using the Bag of Words model6.1.1 Applying the Bag of Words model to movie reviews6.1.2 Cleaning all the movie reviews6.1.3 Exploratory data analysis on your Bag of Words6.2 Building a sentiment classifier using logistic regression6.2.1 Setting up the training for your model6.2.2 Performing the training for your model6.3 Making predictions using your sentiment classifier6.4 Measuring the effectiveness of your classifier6.5 Creating the softmax-regression sentiment classifier6.6 Submitting your results to KaggleSummary
7 Automatically clustering data
7.1 Traversing files in TensorFlow7.2 Extracting features from audio7.3 Using k-means clustering7.4 Segmenting audio7.5 Clustering with a self-organizing map7.6 Applying clusteringSummary
8 Inferring user activity from Android accelerometer data
8.1 The User Activity from Walking dataset8.1.1 Creating the dataset8.1.2 Computing jerk and extracting the feature vector8.2 Clustering similar participants based on jerk magnitudes8.3 Different classes of user activity for a single participantSummary
9 Hidden Markov models
9.1 Example of a not-so-interpretable model9.2 Markov model9.3 Hidden Markov model9.4 Forward algorithm9.5 Viterbi decoding9.6 Uses of HMMs9.6.1 Modeling a video9.6.2 Modeling DNA9.6.3 Modeling an image9.7 Application of HMMsSummary
10 Part-of-speech tagging and word-sense disambiguation
10.1 Review of HMM example: Rainy or Sunny10.2 PoS tagging10.2.1 The big picture: Training and predicting PoS with HMMs10.2.2 Generating the ambiguity PoS tagged dataset10.3 Algorithms for building the HMM for PoS disambiguation10.3.1 Generating the emission probabilities10.4 Running the HMM and evaluating its output10.5 Getting more training data from the Brown Corpus10.6 Defining error bars and metrics for PoS taggingSummary
Part 3 The neural network paradigm
11 A peek into autoencoders
11.1 Neural networks11.2 Autoencoders11.3 Batch training11.4 Working with images11.5 Application of autoencodersSummary
12 Applying autoencoders: The CIFAR-10 image dataset
12.1 What is CIFAR-10?12.1.1 Evaluating your CIFAR-10 autoencoder12.2 Autoencoders as classifiers12.2.1 Using the autoencoder as a classifier via loss12.3 Denoising autoencoders12.4 Stacked deep autoencodersSummary
13 Reinforcement learning
13.1 Formal notions13.1.1 Policy13.1.2 Utility13.2 Applying reinforcement learning13.3 Implementing reinforcement learning13.4 Exploring other applications of reinforcement learningSummary
14 Convolutional neural networks
14.1 Drawback of neural networks14.2 Convolutional neural networks14.3 Preparing the image14.3.1 Generating filters14.3.2 Convolving using filters14.3.3 Max pooling14.4 Implementing a CNN in TensorFlow14.4.1 Measuring performance14.4.2 Training the classifier14.5 Tips and tricks to improve performance14.6 Application of CNNsSummary
15 Building a real-world CNN: VGG -Face and VGG -Face Lite
15.1 Making a real-world CNN architecture for CIFAR-1015.1.1 Loading and preparing the CIFAR-10 image data15.1.2 Performing data augmentation15.2 Building a deeper CNN architecture for CIFAR-1015.2.1 CNN optimizations for increasing learned parameter resilience15.3 Training and applying a better CIFAR-10 CNN15.4 Testing and evaluating your CNN for CIFAR-1015.4.1 CIFAR-10 accuracy results and ROC curves15.4.2 Evaluating the softmax predictions per class15.5 Building VGG -Face for facial recognition15.5.1 Picking a subset of VGG -Face for training VGG -Face Lite15.5.2 TensorFlow’s Dataset API and data augmentation15.5.3 Creating a TensorFlow dataset15.5.4 Training using TensorFlow datasets15.5.5 VGG -Face Lite model and training15.5.6 Training and evaluating VGG -Face Lite15.5.7 Evaluating and predicting with VGG -Face LiteSummary
16 Recurrent neural networks
16.1 Introduction to RNNs16.2 Implementing a recurrent neural network16.3 Using a predictive model for time-series data16.4 Applying RNNsSummary
17 LSTMs and automatic speech recognition
17.1 Preparing the LibriSpeech corpus17.1.1 Downloading, cleaning, and preparing LibriSpeech OpenSLR data17.1.2 Converting the audio17.1.3 Generating per-audio transcripts17.1.4 Aggregating audio and transcripts17.2 Using the deep-speech model17.2.1 Preparing the input audio data for deep speech17.2.2 Preparing the text transcripts as character-level numerical data17.2.3 The deep-speech model in TensorFlow17.2.4 Connectionist temporal classification in TensorFlow17.3 Training and evaluating deep speechSummary
18 Sequence-to-sequence models for chatbots
18.1 Building on classification and RNNs18.2 Understanding seq2seq architecture18.3 Vector representation of symbols18.4 Putting it all together18.5 Gathering dialogue dataSummary
19 Utility landscape
19.1 Preference model19.2 Image embedding19.3 Ranking imagesSummaryWhat’s next
appendix Installation instructions
A.1 Installing the book’s code with DockerA.1.1 Installing Docker in WindowsA.1.2 Installing Docker in LinuxA.1.3 Installing Docker in macOSA.1.4 Using DockerA.2 Getting the data and storing modelsA.3 Necessary librariesA.4 Converting the call-center example to TensorFlow2A.4.1 The call-center example with TF2
index

Content preview from Machine Learning with TensorFlow, Second Edition

7 Automatically clustering data

This chapter covers

Performing basic clustering with k-means
Representing audio
Segmenting audio
Clustering with a self-organizing map

Suppose that you have a collection of not-pirated, totally legal MP3s on your hard drive. All your songs are crowded into one massive folder. Perhaps automatically grouping similar songs into categories such as Country, Rap, and Rock would help organize them. This act of assigning an item to a group (such as an MP3 to a playlist) in an unsupervised fashion is called clustering.

Chapter 6 assumes that you’re given a training dataset of correctly labeled data. Unfortunately, you don’t always have that luxury when you collect data in the real world. Suppose that you want to divide ...

Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.

Read now

Unlock full access

More than 5,000 organizations count on O’Reilly

O’Reilly covers everything we've got, with content to help us build a world-class technology community, upgrade the capabilities and competencies of our teams, and improve overall team performance as well as their engagement.

Julian F.

Head of Cybersecurity

I wanted to learn C and C++, but it didn't click for me until I picked up an O'Reilly book. When I went on the O’Reilly platform, I was astonished to find all the books there, plus live events and sandboxes so you could play around with the technology.

Addison B.

Field Engineer

I’ve been on the O’Reilly platform for more than eight years. I use a couple of learning platforms, but I'm on O'Reilly more than anybody else. When you're there, you start learning. I'm never disappointed.

Amir M.

Data Platform Tech Lead

I'm always learning. So when I got on to O'Reilly, I was like a kid in a candy store. There are playlists. There are answers. There's on-demand training. It's worth its weight in gold, in terms of what it allows me to do.

Mark W.

Embedded Software Engineer

Publisher Resources

ISBN: 9781617297717Publisher Support Publisher Website

Cloud Computing

Data Engineering

Data Science

AI & ML

Programming Languages

Software Architecture

IT/Ops

Security

Design

Business

Soft Skills

Machine Learning with TensorFlow, Second Edition

by Chris Mattmann

7 Automatically clustering data

Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.