book

Mastering Machine Learning Algorithms - Second Edition

Name: Mastering Machine Learning Algorithms - Second Edition
Author: Giuseppe Bonaccorso
ISBN: 9781838820299

by Giuseppe Bonaccorso

January 2020

Intermediate to advanced

798 pages

20h 28m

English

Packt Publishing

Read now

Unlock full access

Preface
Who this book is forWhat this book coversTo get the most out of this bookDownload the example code filesDownload the color imagesConventions usedGet in touchReviews
Machine Learning Model Fundamentals
Models and dataStructure and properties of the datasetsLimited Sample PopulationsScaling datasetsWhiteningTraining, validation, and test setsCross-validationCharacteristics of a machine learning modelLearnabilityCapacity of a modelVapnik-Chervonenkis capacityBias of an estimatorUnderfittingVariance of an estimatorOverfittingThe Cramér-Rao boundSummaryFurther reading
Loss Functions and Regularization
Defining loss and cost functionsExamples of cost functionsMean squared errorHuber cost functionHinge cost functionCategorical cross-entropyRegularizationExamples of Regularization TechniquesL2 or Ridge regularizationL1 or Lasso regularizationElasticNetEarly stoppingSummaryFurther reading
Introduction to Semi-Supervised Learning
Semi-supervised scenarioCausal scenariosTransductive learningInductive learningSemi-supervised assumptionsSmoothness assumptionCluster assumptionManifold assumptionGenerative Gaussian MixtureGenerative Gaussian Mixture theoryExample of a Generative Gaussian MixtureGenerative Gaussian Mixtures summaryWeighted log-likelihoodSelf-TrainingSelf-Training theoryExample of Self-Training with the Iris datasetSelf-Training summaryCo-TrainingCo-Training theoryExample of Co-Training with the Wine datasetCo-Training summarySummaryFurther reading
Advanced Semi-Supervised Classification
Contrastive Pessimistic Likelihood EstimationCPLE TheoryExample of contrastive pessimistic likelihood estimationCPLE SummarySemi-supervised Support Vector Machines (S3VM)S3VM TheoryExample of S3VMS3VM SummaryTransductive Support Vector Machines (TSVM)TSVM TheoryExample of TSVMAnalysis of different TSVM configurationsTSVM SummarySummaryFurther reading
Graph-Based Semi-Supervised Learning
Label propagationExample of label propagationLabel propagation in scikit-learnLabel spreadingExample of label spreadingIncreasing the smoothness with Laplacian regularizationLabel propagation based on Markov random walksExample of label propagation based on Markov random walksManifold learningIsomapExample of IsomapLocally linear embeddingExample of LLELaplacian Spectral EmbeddingExample of Laplacian Spectral Embeddingt-SNEExample of t-distributed stochastic neighbor embeddingSummaryFurther reading
Clustering and Unsupervised Models
K-nearest neighborsK-d treesBall treesFitting a KNN modelExample of KNN with scikit-learnK-meansK-means++Example of K-means with scikit-learnEvaluation metricsHomogeneity scoreCompleteness scoreAdjusted Rand indexSilhouette scoreSummaryFurther reading
Advanced Clustering and Unsupervised Models
Fuzzy C-meansExample of Fuzzy C-means with SciKit-FuzzySpectral clusteringExample of spectral clustering with scikit-learnDBSCANExample of DBSCAN with scikit-learnThe Calinski-Harabasz scoreThe Davies-Bouldin scoreAnalysis of DBSCAN resultsSummaryFurther reading
Clustering and Unsupervised Models for Marketing
BiclusteringExample of Spectral Biclustering with Scikit-LearnIntroduction to Market Basket Analysis with the Apriori AlgorithmExample of Apriori in PythonSummaryFurther reading
Generalized Linear Models and Regression
GLMsLeast Squares EstimationBias and Variance of Least Squares EstimatorsExample of Linear regression with PythonComputing Linear regression Confidence Intervals with StatsmodelsIncreasing the robustness to outliers with Huber lossOther regression techniquesRidge RegressionExample of Ridge Regression with scikit-learnRisk modeling with Lasso and Logistic RegressionExample of Risk modeling with Lasso and Logistic RegressionPolynomial RegressionExamples of Polynomial RegressionsIsotonic RegressionExample of Isotonic RegressionSummaryFurther reading

Introduction to Time-Series Analysis
Time-seriesSmoothingIntroduction to linear models for time-seriesAutocorrelationAR, MA, and ARMA processesModeling non-stationary trend models with ARIMASummaryFurther reading
Bayesian Networks and Hidden Markov Models
Conditional probabilities and Bayes' theoremConjugate priorsBayesian networksSampling from a Bayesian networkDirect samplingA gentle introduction to Markov ChainsGibbs samplingThe Metropolis-Hastings algorithmSampling using PyMC3Running the Sampling ProcessSampling using PyStanHidden Markov ModelsThe Forward-Backward algorithmForward phaseBackward phaseHMM parameter estimationThe Viterbi algorithmFinding the most likely hidden state sequence using the Viterbi algorithm and hmmlearnSummaryFurther reading
The EM Algorithm
MLE and MAP LearningEM AlgorithmConvex functions and the Jensen's inequalityApplication of the Jensen's inequality to the EM algorithmAn example of parameter estimationGaussian MixtureExample of Gaussian Mixture with scikit-learnDetermining the optimal number of components using AIC and BICAutomatic component selection using Bayesian Gaussian MixtureSummaryFurther reading
Component Analysis and Dimensionality Reduction
Factor AnalysisLinear relation analysisExample of Factor Analysis with scikit-learnPrincipal Component AnalysisComponent importance evaluationExample of PCA with scikit-learnKernel PCASparse PCAIndependent Component AnalysisExample of FastICA with scikit-learnAddendum to Hidden Markov ModelsSummaryFurther reading
Hebbian Learning
Hebb's ruleAnalysis of the Covariance RuleExample of application of the covariance ruleWeight vector stabilization and Oja's ruleSanger's networkExample of Sanger's networkRubner-Tavan's networkExample of Rubner-Tavan's NetworkSelf-organizing mapsKohonen MapsExample of SOMSummaryFurther reading
Fundamentals of Ensemble Learning
Ensemble learning fundamentalsRandom forestsRandom forest fundamentalsWhy use Decision Trees?Random forests and the bias-variance trade-offExample of random forest with scikit-learnFeature importanceAdaBoostAdaBoost.SAMMEAdaBoost.SAMME.RAdaBoost.R2Example of AdaBoost with scikit-learnSummaryFurther reading
Advanced Boosting Algorithms
Gradient boostingLoss functions for gradient boostingExample of gradient tree boosting with scikit-learnExample of gradient boosting with XGBoostEvaluating the predictive power of the featuresEnsembles of voting classifiersExample of voting classifiers with scikit-learnEnsemble learning as model selectionSummaryFurther reading
Modeling Neural Networks
The basic artificial neuronThe perceptronExample of a Perceptron with scikit-learnMultilayer Perceptrons (MLPs)Activation functionsSigmoid and Hyperbolic TangentRectifier activation functionsSoftmaxThe back-propagation algorithmStochastic gradient descent (SGD)Weight initializationExample of MLP with TensorFlow and KerasSummaryFurther reading
Optimizing Neural Networks
Optimization algorithmsGradient perturbationMomentum and Nesterov momentumSGD with Momentum in TensorFlow and KerasRMSPropRMSProp in TensorFlow and KerasAdamAdam in TensorFlow and KerasAdaGradAdaGrad with TensorFlow and KerasAdaDeltaAdaDelta in TensorFlow and KerasRegularization and DropoutRegularizationRegularization in TensorFlow and KerasDropoutDropout with TensorFlow and KerasBatch normalizationExample of batch normalization with TensorFlow and KerasSummaryFurther reading
Deep Convolutional Networks
Deep convolutional networksConvolutional operatorsBidimensional discrete convolutionsStrides and PaddingAtrous convolutionSeparable convolutionTranspose convolutionPooling layersOther helpful layersExample of a deep convolutional network with TensorFlow and KerasExample of a deep convolutional network with TensorFlow/Keras and data augmentationSummaryFurther reading
Recurrent Neural Networks
Recurrent networksBackpropagation through timeLimitations of BPTTLong Short-Term Memory (LSTM)Gated Recurrent Unit (GRU)Example of an LSTM with TensorFlow and KerasTransfer learningSummaryFurther reading
Autoencoders
AutoencodersExample of a deep convolutional autoencoder with TensorFlowDenoising autoencodersExample of a denoising autoencoder with TensorFlowSparse autoencodersAdding sparseness to the Fashion MNIST deep convolutional autoencoderVariational autoencodersExample of a VAE with TensorFlowSummaryFurther reading
Introduction to Generative Adversarial Networks
Adversarial trainingDeep Convolutional GANsExample of DCGAN with TensorFlowMode collapseWasserstein GANExample of WGAN with TensorFlowSummaryFurther reading
Deep Belief Networks
Introduction to Markov random fieldsRestricted Boltzmann MachinesContrastive DivergenceDeep Belief NetworksExample of an unsupervised DBN in PythonExample of a supervised DBN in PythonSummaryFurther reading
Introduction to Reinforcement Learning
Fundamental concepts of RLThe Markov Decision ProcessEnvironmentRewardsA checkerboard environment in PythonPolicyPolicy iterationPolicy iteration in the checkerboard environmentValue iterationValue iteration in the checkerboard environmentThe TD(0) algorithmTD(0) in the checkerboard environmentSummaryFurther reading
Advanced Policy Estimation Algorithms
TD(λ) algorithmTD(λ) in a more complex checkerboard environmentActor-Critic TD(0) in the checkerboard environmentSARSA algorithmSARSA in the checkerboard environmentQ-learningQ-learning in the checkerboard environmentQ-learning modeling the policy with a neural networkDirect policy search through policy gradientExample of policy gradient with OpenAI Gym CartpoleSummaryFurther reading
Other Books You May Enjoy
Index

Overview

Dive into the intricate world of machine learning with 'Mastering Machine Learning Algorithms, Second Edition.' This expanded guide walks you through implementation, theoretical foundation, and advanced techniques in the field, ensuring you can tackle complex ML problems. By mastering these algorithms, you will gain the skills to handle diverse data challenges using Python, TensorFlow, and more.

What this Book will help me do

Understand and characterize various machine learning algorithms, delving deep into their mechanisms and appropriate applications.
Implement advanced techniques in supervised, semi-supervised, unsupervised learning, and reinforcement learning using Python libraries.
Master time-series analysis, regression models, and their applications in real-world forecasting and decision-making scenarios.
Develop skills to train, optimize, and evaluate artificial neural network models, including deep learning architectures.
Apply contemporary techniques such as autoencoders, generative adversarial networks (GANs), and probabilistic models to solve cutting-edge challenges.

Author(s)

Giuseppe Bonaccorso is a machine learning expert, bringing years of industry experience in data science and software engineering. Giuseppe has a pragmatic teaching style, focusing on explaining the core principles and methods of ML algorithms while bridging the gap between theory and practice. His ability to demystify complex topics has made his past books highly revered among technical learners.

Who is it for?

This book is designed for data professionals looking to elevate their understanding of machine learning algorithms. If you're a data scientist, software engineer, or research enthusiast with fundamental Python knowledge, this book is tailored for you. It connects theory with practical implementations, appealing to those keen to master algorithm foundations and applications in ML.

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: 9781838820299

Cloud Computing

Data Engineering

Data Science

AI & ML

Programming Languages

Software Architecture

IT/Ops

Security

Design

Business

Soft Skills