All you need to know about Machine Learning in a hundred pages
Supervised and unsupervised learning, support vector machines, neural networks, ensemble methods, gradient descent, cluster analysis and dimensionality reduction, autoencoders and transfer learning, feature engineering and hyperparameter tuning! Math, intuition, illustrations, all in just a hundred pages!
Author(s): Andriy Burkov
Publisher: Andriy Burkov
Year: 2019
Language: English
Pages: 160
Foreword
Preface
Who This Book is For
How to Use This Book
Should You Buy This Book?
Introduction
What is Machine Learning
Types of Learning
Supervised Learning
Unsupervised Learning
Semi-Supervised Learning
Reinforcement Learning
How Supervised Learning Works
Why the Model Works on New Data
Notation and Definitions
Notation
Data Structures
Capital Sigma Notation
Capital Pi Notation
Operations on Sets
Operations on Vectors
Functions
Max and Arg Max
Assignment Operator
Derivative and Gradient
Random Variable
Unbiased Estimators
Bayes' Rule
Parameter Estimation
Parameters vs. Hyperparameters
Classification vs. Regression
Model-Based vs. Instance-Based Learning
Shallow vs. Deep Learning
Fundamental Algorithms
Linear Regression
Problem Statement
Solution
Logistic Regression
Problem Statement
Solution
Decision Tree Learning
Problem Statement
Solution
Support Vector Machine
Dealing with Noise
Dealing with Inherent Non-Linearity
k-Nearest Neighbors
Anatomy of a Learning Algorithm
Building Blocks of a Learning Algorithm
Gradient Descent
How Machine Learning Engineers Work
Learning Algorithms' Particularities
Basic Practice
Feature Engineering
One-Hot Encoding
Binning
Normalization
Standardization
Dealing with Missing Features
Data Imputation Techniques
Learning Algorithm Selection
Three Sets
Underfitting and Overfitting
Regularization
Model Performance Assessment
Confusion Matrix
Precision/Recall
Accuracy
Cost-Sensitive Accuracy
Area under the ROC Curve (AUC)
Hyperparameter Tuning
Cross-Validation
Neural Networks and Deep Learning
Neural Networks
Multilayer Perceptron Example
Feed-Forward Neural Network Architecture
Deep Learning
Convolutional Neural Network
Recurrent Neural Network
Problems and Solutions
Kernel Regression
Multiclass Classification
One-Class Classification
Multi-Label Classification
Ensemble Learning
Boosting and Bagging
Random Forest
Gradient Boosting
Learning to Label Sequences
Sequence-to-Sequence Learning
Active Learning
Semi-Supervised Learning
One-Shot Learning
Zero-Shot Learning
Advanced Practice
Handling Imbalanced Datasets
Combining Models
Training Neural Networks
Advanced Regularization
Handling Multiple Inputs
Handling Multiple Outputs
Transfer Learning
Algorithmic Efficiency
Unsupervised Learning
Density Estimation
Clustering
K-Means
DBSCAN and HDBSCAN
Determining the Number of Clusters
Other Clustering Algorithms
Dimensionality Reduction
Principal Component Analysis
UMAP
Outlier Detection
Other Forms of Learning
Metric Learning
Learning to Rank
Learning to Recommend
Factorization Machines
Denoising Autoencoders
Self-Supervised Learning: Word Embeddings
Conclusion
What Wasn't Covered
Topic Modeling
Gaussian Processes
Generalized Linear Models
Probabilistic Graphical Models
Markov Chain Monte Carlo
Generative Adversarial Networks
Genetic Algorithms
Reinforcement Learning
Acknowledgements
Index
