Chapter 1, A Gentle Introduction to Machine Learning, introduces the world of machine learning, explaining the fundamental concepts of the most important approaches to creating intelligent applications and focusing on the different kinds of learning methods.
Chapter 2, Important Elements in Machine Learning, explains the mathematical concepts regarding the most common machine learning problems, including the concept of learnability and some important elements of information theory. This chapter contains theoretical elements, but it's extremely helpful if you are learning this topic from scratch because it provides an insight into the most important mathematical tools employed in the majority of algorithms.
Chapter 3, Feature Selection and Feature Engineering, describes the most important techniques for preprocessing a dataset, selecting the most informative features, and reducing the original dimensionality.
Chapter 4, Regression Algorithms, describes the linear regression algorithm and its optimizations: Ridge, Lasso, and ElasticNet. It continues with more advanced models that can be employed to solve non-linear regression problems or to mitigate the effect of outliers.
Chapter 5, Linear Classification Algorithms, introduces the concept of linear classification, focusing on logistic regression, perceptrons, stochastic gradient descent algorithms, and passive-aggressive algorithms. The second part of the chapter covers the most important evaluation metrics, which are used to measure the performance of a model and find the optimal hyperparameter set.
Chapter 6, Naive Bayes and Discriminant Analysis, explains the Bayes probability theory and describes the structure of the most diffused Naive Bayes classifiers. In the second part, linear and quadratic discriminant analysis is analyzed with some concrete examples.
Chapter 7, Support Vector Machines, introduces the SVM family of algorithms, focusing on both linear and non-linear classification problems thanks to the employment of the kernel trick. The last part of the chapter covers support vector regression and more complex classification models.
Chapter 8, Decision Trees and Ensemble Learning, explains the concept of a hierarchical decision process and describes the concepts of decision tree classification, random forests, bootstrapped and bagged trees, and voting classifiers.
Chapter 9, Clustering Fundamentals, introduces the concept of clustering, describing the Gaussian mixture, K-Nearest Neighbors, and K-means algorithms. The last part of the chapter covers different approaches to determining the optimal number of clusters and measuring the performance of a model.
Chapter 10, Advanced Clustering, introduces more complex clustering techniques (DBSCAN, Spectral Clustering, and Biclustering) that can be employed when the dataset structure is non-convex. In the second part of the chapter, two online clustering algorithms (mini-batch K-means and BIRCH) are introduced.
Chapter 11, Hierarchical Clustering, continues the explanation of more complex clustering algorithms started in the previous chapter and introduces the concepts of agglomerative clustering and dendrograms.
Chapter 12, Introducing Recommendation Systems, explains the most diffused algorithms employed in recommender systems: content- and user-based strategies, collaborative filtering, and alternating least square. A complete example based on Apache Spark shows how to process very large datasets using the ALS algorithm.
Chapter 13, Introduction to Natural Language Processing, explains the concept of the Bag-of-Words strategy and introduces the most important techniques required to efficiently process natural language datasets (tokenizing, stemming, stop-word removal, tagging, and vectorizing). An example of a classifier based on the Reuters dataset is also discussed in the last part of the chapter.
Chapter 14, Topic Modeling and Sentiment Analysis in NLP, introduces the concept of topic modeling and describes the most important algorithms, such as latent semantic analysis (both deterministic and probabilistic) and latent Dirichlet allocation. The second part of the chapter covers the problem of word embedding and sentiment analysis, explaining the most diffused approaches to address it.
Chapter 15, Introducing Neural Networks, introduces the world of deep learning, explaining the concept of neural networks and computational graphs. In the second part of the chapter, the high-level deep learning framework Keras is presented with a concrete example of a Multi-layer Perceptron.
Chapter 16, Advanced Deep Learning Models, explains the basic functionalities of the most important deep learning layers, with Keras examples of deep convolutional networks and recurrent (LSTM) networks for time-series processing. In the second part of the chapter, the TensorFlow framework is briefly introduced, along with some examples that expose some of its basic functionalities.
Chapter 17, Creating a Machine Learning Architecture, explains how to define a complete machine learning pipeline, focusing on the peculiarities and drawbacks of each step.
