Machine and Deep Learning Using MATLAB E-Book

Kamal I. M. Al-Malah

• Herausgeber: John Wiley & Sons
• Kategorie: Fachliteratur
• Sprache: Englisch

Seitenzahl: 704

Veröffentlichungsjahr: 2023

Beschreibung

MACHINE AND DEEP LEARNING In-depth resource covering machine and deep learning methods using MATLAB tools and algorithms, providing insights and algorithmic decision-making processes Machine and Deep Learning Using MATLAB introduces early career professionals to the power of MATLAB to explore machine and deep learning applications by explaining the relevant MATLAB tool or app and how it is used for a given method or a collection of methods. Its properties, in terms of input and output arguments, are explained, the limitations or applicability is indicated via an accompanied text or a table, and a complete running example is shown with all needed MATLAB command prompt code. The text also presents the results, in the form of figures or tables, in parallel with the given MATLAB code, and the MATLAB written code can be later used as a template for trying to solve new cases or datasets. Throughout, the text features worked examples in each chapter for self-study with an accompanying website providing solutions and coding samples. Highlighted notes draw the attention of the user to critical points or issues. Readers will also find information on: * Numeric data acquisition and analysis in the form of applying computational algorithms to predict the numeric data patterns (clustering or unsupervised learning) * Relationships between predictors and response variable (supervised), categorically sub-divided into classification (discrete response) and regression (continuous response) * Image acquisition and analysis in the form of applying one of neural networks, and estimating net accuracy, net loss, and/or RMSE for the successive training, validation, and testing steps * Retraining and creation for image labeling, object identification, regression classification, and text recognition Machine and Deep Learning Using MATLAB is a useful and highly comprehensive resource on the subject for professionals, advanced students, and researchers who have some familiarity with MATLAB and are situated in engineering and scientific fields, who wish to gain mastery over the software and its numerous applications.

Leseprobe

Machine and Deep Learning Using MATLAB

Algorithms and Tools for Scientists and Engineers

Kamal I. M. Al-Malah

This edition first published 2024

© 2024 Kamal I. M. Al-Malah

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by law. Advice on how to obtain permission to reuse material from this title is available at http://www.wiley.com/go/permissions.

The right of Kamal I. M. Al-Malah to be identified as the author of this work has been asserted in accordance with law.

Registered Office

John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA

For details of our global editorial offices, customer services, and more information about Wiley products visit us at www.wiley.com.

Wiley also publishes its books in a variety of electronic formats and by print-on-demand. Some content that appears in standard print versions of this book may not be available in other formats.

Limit of Liability/Disclaimer of Warranty

MATLAB® is a trademark of The MathWorks, Inc. and is used with permission. The MathWorks does not warrant the accuracy of the text or exercises in this book. This work’s use or discussion of MATLAB® software or related products does not constitute endorsement or sponsorship by The MathWorks of a particular pedagogical approach or particular use of the MATLAB® software. While the publisher and authors have used their best efforts in preparing this work, they make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives, written sales materials or promotional statements for this work. This work is sold with the understanding that the publisher is not engaged in rendering professional services. The advice and strategies contained herein may not be suitable for your situation. You should consult with a specialist where appropriate. The fact that an organization, website, or product is referred to in this work as a citation and/or potential source of further information does not mean that the publisher and authors endorse the information or services the organization, website, or product may provide or recommendations it may make. Further, readers should be aware that websites listed in this work may have changed or disappeared between when this work was written and when it is read. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.

A catalogue record for this book is available from the Library of Congress

Hardback ISBN: 9781394209088; ePub ISBN: 9781394209101; ePDF ISBN: 9781394209095; oBook ISBN: 9781394209118

Cover image: [insert based on cover mechanical]

Cover Design: Wiley

Set in 9.5/12.5pt STIXTwoText by Integra Software Services Pvt. Ltd, Pondicherry, India

Preface

Welcome to “Machine and Deep Learning Using MATLAB Algorithms and Tools for Scientists and Engineers.” In today’s data-driven world, machine learning and deep learning have become indispensable tools for scientists and engineers across various disciplines. This book aims to provide a comprehensive guide to understanding and applying these techniques using MATLAB algorithms and tools. Divided into ten chapters, “Machine and Deep Learning Using MATLAB Algorithms and Tools for Scientists and Engineers” offers a comprehensive coverage of both machine learning and deep learning techniques. The book takes a step-by-step approach, guiding readers through the process of acquiring, analyzing, and predicting patterns in both numeric and image data.

The first five chapters provide a solid foundation in machine learning, covering unsupervised learning, classification, predictive model improvement, linear regression, and neural networks. Through clear explanations, practical examples, and hands-on case studies, readers will develop the knowledge and skills necessary to apply these techniques to their own scientific and engineering endeavors. Readers will delve into various techniques that are widely used in the field, including clustering, classification, regression, and feature selection. These five chapters provide a solid foundation in machine learning concepts and methods, allowing readers to gain a deep understanding of how to apply these techniques to real-world problems.

Chapter One focuses on unsupervised machine learning techniques. It explores methods such as Classical Multidimensional Scaling, Principal Component Analysis (PCA), k-Means Clustering, Gaussian Mixture Model (GMM) Clustering, and Feature Selection Using Laplacian. The chapter provides tools for visualization and observation of clusters, as well as Hierarchical Clustering. Readers are guided through practical case studies, including analyzing Iris Flower Features Data, Ionosphere Data Features, Small Car Data, and Seeds Features Data.

Chapter Two delves into fitting data using different classification models. It introduces classification techniques like K-Nearest Neighbors (KNN), Binary Decision Tree, Naïve Bayes, Discriminant Analysis (DA), Support Vector Machine (SVM), Multiclass SVM, and Binary Linear Classifier. The chapter also showcases the MATLAB Classification Learner app, which allows users to explore these techniques without writing code. Through case studies involving Mushroom Edibility Data, Adult Census Income Data, White Wine Classification, Cardiac Arrhythmia Data, and Breast Cancer Diagnosis Data, readers gain practical experience in implementing the discussed methods.

Chapter Three covers methods for improving predictive models in machine learning. It explores topics such as Cross Validation, Feature Transformation and Selection, Factor Analysis, Sequential Feature Selection (SFS), Dummy Variables, and Ensemble Learning. The chapter also introduces Feature Selection Using Neighborhood Component Analysis (NCA) for regression problems. Readers will learn how to utilize the Regression Learner app to perform feature selection and transformation. Through case studies involving Ionosphere Data, Sonar Dataset, White Wine Classification, and Small Car Data (Regression Case), readers gain hands-on experience in implementing these techniques.

Chapter Four focuses on ML linear regression methods. It covers various approaches, including fitting Linear Regression Models using the fitlm/fitglm function, Non-Parametric Regression Models, Gaussian Process Regression (GPR), Regularized Parametric Linear Regression, Lasso, and Stepwise Parametric Linear Regression. Practical case studies, such as Boston House Price, The Forest Fires Data, The Parkinson’s Disease Telemonitoring Data, and The Car Fuel Economy Data, enable readers to apply these methods to real-world datasets.

Chapter Five explores neural networks for classification and regression tasks. It introduces Feed-Forward Neural Networks and the Neural Network Pattern Recognition (nprtool) tool for classification problems, as well as Feed-Forward Neural Network Regression and the Neural Network Regression (nftool) for data fitting. The chapter covers training the Neural Network Regression Model using the fitrnet function, finding the optimum regularization strength through cross-validation, and custom hyperparameter optimization in neural network regression. Case studies involving Mushroom Edibility Data, 1994 Adult Census Income Data, Breast Cancer Diagnosis, Small Car Data (Regression Case), and Boston House Price enable readers to implement these techniques effectively.

The remaining five chapters focus on the exciting world of deep learning. Deep learning has gained significant popularity due to its ability to analyze complex patterns in large datasets, particularly in the field of image analysis. In these chapters, readers will explore topics such as neural networks, transfer learning, convolutional neural networks (CNNs), object detection, and recurrent neural networks (RNNs). Additionally, the book introduces image/video-based MATLAB tools that facilitate the implementation of deep learning algorithms and enable the analysis of images and videos. These remaining five chapters build upon the foundation established in the earlier chapters, delving into more advanced topics and techniques. Through practical examples, case studies, and hands-on implementation, readers acquire the knowledge and skills necessary to apply transfer learning, CNN architecture, object detection, RNNs, and image/video-based MATLAB tools to a variety of real-world problems.

Chapter Six focuses on transfer learning of pre-trained neural networks. It covers topics such as Data Stores in MATLAB, Image and Augmented Image Datastores, Retraining for Image Recognition, Convolutional Neural Network (CNN) Layers, Channels and Activations, and Feature Extraction for Machine Learning. Additionally, the chapter explores network object prediction explainers like Occlusion Sensitivity, imageLIME Features Explainer, and gradCAM Features Explainer. Through case studies involving CNN retraining for various predictions, such as Round Worms Alive or Dead, Food Images, Merchandise Data, Musical Instrument Spectrograms, and Fruit/Vegetable Varieties, readers gain practical experience in applying these techniques.

Chapter Seven dives into the architecture and training of Convolutional Neural Networks (CNNs). It covers topics like Training Options, Filters in Convolution Layers, Validation Data, Improving Network Performance, Image Augmentation using the Flowers Dataset, Directed Acyclic Graphs Networks, Deep Network Designer (DND), and Semantic Segmentation. Readers are presented with case studies that involve creating CNNs for predictions such as Round Worms Alive or Dead, Food Images, Merchandise Data, Musical Instrument Spectrograms, and Chest X-ray. By following these examples, readers develop a deeper understanding of CNN architecture and training methodologies.

Chapter Eight explores regression classification and object detection. It covers topics like Preparing Data for Regression, Deep Network Designer (DND) for Regression, YOLO Object Detectors, Object Detection Using Regions with Convolutional Neural Networks (R-CNN), R-CNN Transfer Learning (Re-Training), evaluateDetectionPrecision Function for Precision Metric, and evaluateDetectionMissRate for Miss Rate Metric. Through case studies involving testing object detectors, creating CNN-based object detectors, and creating GoogleNet-Based Fast R-CNN object detectors, readers gain hands-on experience in regression classification and object detection tasks.

Chapter Nine focuses on Recurrent Neural Networks (RNNs). It covers topics such as Long Short-Term Memory (LSTM) and BiLSTM Networks, Classifying Categorical Sequences, Sequence-to-Sequence Regression using Deep Learning, Classifying Text Data for Factory Equipment Failure Analysis, Word-By-Word Text Generation, Training Networks for Time Series Forecasting using Deep Network Designer (DND), and Network Training with Numeric Features. The chapter provides case studies involving text classification, text regression, and multivariate classification, allowing readers to apply RNN techniques to real-world datasets.

Chapter Ten covers image/video-based MATLAB tools. It introduces tools such as Image Labeler (IL), Video Labeler (VL), Ground Truth Labeler (GTL), Experiment Manager (EM), and Image Batch Processor (IBP). Readers are guided through case studies involving tasks like video labeling, training, and prediction, as well as hyperparameter tuning for CNN retraining. By working on these examples, readers gain hands-on experience with the image/video-based MATLAB tools and learn how to apply them effectively.

By covering both machine learning and deep learning, this book provides readers with a comprehensive understanding of these powerful techniques. The step-by-step approach ensures that readers can gradually build their knowledge and skills, starting with the fundamentals of machine learning and progressing to more advanced topics in deep learning. Throughout the book, readers will gain hands-on experience through practical examples and case studies, enabling them to apply the learned techniques to their own projects and research.

Each chapter follows a consistent method of approach to ensure clarity and ease of understanding. Firstly, the MATLAB built-in functions relevant to the topic are introduced, along with their properties, limitations, and applicability. Complete running examples are provided, along with the corresponding MATLAB code. Results in the form of figures and tables are presented alongside the code, enabling a better grasp of the concepts. Additionally, MATLAB tools and apps are explained and utilized where applicable, providing an alternative approach to achieving results. Quizzes are included in each chapter to test your understanding, and important notes are highlighted to draw your attention to critical points and potential pitfalls. Finally, end-of-chapter problems are tailored to reinforce the concepts covered, serving as opportunities to apply the methods to real-world case studies.

Throughout the book, a wide range of datasets are used to illustrate the concepts and techniques discussed. From the Iris Flower Features Data to the Boston House Price dataset, from Mushroom Edibility Data to Factory Equipment Failure Text Analysis, you will gain hands-on experience in solving various problems using MATLAB algorithms and tools.

It is worth mentioning that the solution manual for the end-of-chapter problems will be available exclusively for instructors, allowing them to guide students through the learning process. However, data sets, images, or videos for all running examples and for those of end of chapter problems will be available at Wiley’s Web companion for all book users and readers. The Companion Website address is: www.wiley.com/go/al-malah/machinelearningmatlab

This book aims to be a valuable resource for anyone seeking to harness the power of machine learning and deep learning in their research or professional endeavors. Whether you are a scientist, engineer, or student, this book equips you with the knowledge and tools necessary to tackle complex data analysis and prediction tasks. The comprehensive coverage and practical approach make this book an invaluable resource for anyone looking to leverage the power of machine learning and deep learning in their work.

We hope you find this book enlightening and empowering as you embark on your journey into the world of machine and deep learning using MATLAB algorithms and tools.

Happy learning!

Kamal Al-Malah

About the Companion Website

This book is accompanied by a companion website which includes a number of resources created by author for students and instructors that you will find helpful.

www.wiley.com/go/al-malah/machinelearningmatlab

The Instructor website includes:

End of Chapter Solutions

The website includes the following resources for each chapter:

Auxiliary Data

Please note that the resources in instructor website are password protected and can only be accessed by instructors who buy the book.

1 Unsupervised Machine Learning (ML) Techniques

Introduction

As quoted by MATLAB R2021b built-in help, machine learning (ML) teaches computers to do what comes naturally to humans: Learn from experience. Machine learning algorithms utilize computational methods to directly learn (or extract) information from data without relying on a deterministic model. The set of algorithms adaptively improve their performance as the number of samples available for learning increases. ML uses two types of learning techniques: Unsupervised and supervised.

The unsupervised