Machine learning : a first course for engineers and scientists

"This book introduces machine learning for readers with some background in basic linear algebra, statistics, probability, and programming. In a coherent statistical framework it covers a selection of supervised machine learning methods, from the most fundamental (k-NN, decision trees, linear and logistic regression) to more advanced methods (deep neural networks, support vector machines, Gaussian processes, random forests and boosting), plus commonly-used unsupervised methods (generative modeling, k-means, PCA, autoencoders and generative adversarial networks). Careful explanations and pseudo-code are presented for all methods. The authors maintain a focus on the fundamentals by drawing connections between methods and discussing general concepts such as loss functions, maximum likelihood, the bias-variance decomposition, ensemble averaging, kernels and the Bayesian approach along with generally useful tools such as regularization, cross validation, evaluation metrics and optimization methods. The final chapters offer practical advice for solving real-world supervised machine learning problems and on ethical aspects of modern machine learning"-- Provided by publisher Cover Half-title Title page Copyright information Contents Acknowledgements Notation 1 Introduction 1.1 Machine Learning Exemplified 1.2 About This Book 1.3 Further Reading 2 Supervised Learning: A First Approach 2.1 Supervised Machine Learning 2.2 A Distance-Based Method: k-NN 2.3 A Rule-Based Method: Decision Trees 2.4 Further Reading 3 Basic Parametric Models and a Statistical Perspective on Learning 3.1 Linear Regression 3.2 Classification and Logistic Regression 3.3 Polynomial Regression and Regularisation 3.4 Generalised Linear Models 3.5 Further Reading 3.A Derivation of the Normal Equations 4 Understanding, Evaluating, and Improving Performance 4.1 Expected New Data Error E[sub(new)]: Performance in Production 4.2 Estimating E[sub(new)] 4.3 The Training Error–Generalisation Gap Decomposition of E[sub(new)] 4.4 The Bias–Variance Decomposition of E[sub(new)] 4.5 Additional Tools for Evaluating Binary Classifiers 4.6 Further Reading 5 Learning Parametric Models 5.1 Principles of Parametric Modelling 5.2 Loss Functions and Likelihood-Based Models 5.3 Regularisation 5.4 Parameter Optimisation 5.5 Optimisation with Large Datasets 5.6 Hyperparameter Optimisation 5.7 Further Reading 6 Neural Networks and Deep Learning 6.1 The Neural Network Model 6.2 Training a Neural Network 6.3 Convolutional Neural Networks 6.4 Dropout 6.5 Further Reading 6.A Derivation of the Backpropagation Equations 7 Ensemble Methods: Bagging and Boosting 7.1 Bagging 7.2 Random Forests 7.3 Boosting and AdaBoost 7.4 Gradient Boosting 7.5 Further Reading 8 Non-linear Input Transformations and Kernels 8.1 Creating Features by Non-linear Input Transformations 8.2 Kernel Ridge Regression 8.3 Support Vector Regression 8.4 Kernel Theory 8.5 Support Vector Classification 8.6 Further Reading 8.A The Representer Theorem 8.B Derivation of Support Vector Classification 9 The Bayesian Approach and Gaussian Processes 9.1 The Bayesian Idea 9.2 Bayesian Linear Regression 9.3 The Gaussian Process 9.4 Practical Aspects of the Gaussian Process 9.5 Other Bayesian Methods in Machine Learning 9.6 Further Reading 9.A The Multivariate Gaussian Distribution 10 Generative Models and Learning from Unlabelled Data 10.1 The Gaussian Mixture Model and Discriminant Analysis 10.2 Cluster Analysis 10.3 Deep Generative Models 10.4 Representation Learning and Dimensionality Reduction 10.5 Further Reading 11 User Aspects of Machine Learning 11.1 Defining the Machine Learning Problem 11.2 Improving a Machine Learning Model 11.3 What If We Cannot Collect More Data? 11.4 Practical Data Issues 11.5 Can I Trust my Machine Learning Model? 11.6 Further Reading 12 Ethics in Machine Learning 12.1 Fairness and Error Functions 12.2 Misleading Claims about Performance 12.3 Limitations of Training Data 12.4 Further Reading Bibliography Index