Terms
Glossary
SohoBook machine learning glossary: key terms, related topics and chapter links.
Affine Transformation
A linear transformation with an added bias; a basic form of many predictive models.
weight
bias
linear function
Introduction
Python Basics
Basic Mathematical Computations
Vector
An ordered set of numbers; in machine learning it often describes an object through features.
matrix
norm
dot product
Introduction
Python Basics
Basic Mathematical Computations
Matrix
A table of numbers used to store datasets and describe linear transformations.
linear algebra
tensor
multiplication
Introduction
Python Basics
Basic Mathematical Computations
Machine Learning
An approach to building systems that find patterns in data and use them for prediction or decision-making.
data
model
training
Introduction
Python Basics
Basic Mathematical Computations
Norm
A numerical measure of vector length or error magnitude, useful for comparing distances and deviations.
vector
distance
error
Introduction
Python Basics
Basic Mathematical Computations
Perceptron
A simple linear classifier that updates weights using errors on training examples.
SVM
decision boundary
weight
Introduction
Prediction Function and Universal Prediction Theory
Perceptron and Support Vector Methods (SVM)
Feature
A measurable characteristic of an object that is supplied to a model as input.
vector
data
scaling
Introduction
Basic Mathematical Computations
Learning Models Based on Statistics and Probability
Entropy
A measure of distribution uncertainty; the more uniform the outcome probabilities, the higher the entropy.
probability
KL Divergence
cross-entropy
Learning Models Based on Statistics and Probability
AI Ethics
A set of questions about risks, fairness, privacy, responsibility, and control in the use of AI.
AGI
risks
responsibility
Prediction Function and Universal Prediction Theory
Artificial Intelligence: Definition, History, and Future
Main Approaches and Ethical Considerations in Machine Learning
Decision Tree
A model that recursively divides feature space using if-then style rules.
classification
entropy
tree
Learning Models Based on Statistics and Probability
Perceptron and Support Vector Methods (SVM)
KL Divergence
A measure of how one probability distribution differs from another.
entropy
distribution
probability
Learning Models Based on Statistics and Probability
Perceptron and Support Vector Methods (SVM)
KNN
A classifier that assigns an object to the class of its nearest known neighbors.
classification
distance
decision boundary
Introduction
Basic Mathematical Computations
Learning Models Based on Statistics and Probability
Naive Bayes
A probabilistic classifier based on Bayes theorem and the assumption of conditional feature independence.
Bayes
probability
classification
Introduction
Basic Mathematical Computations
Learning Models Based on Statistics and Probability
PCA
Principal component analysis, a method that finds new axes of maximum variance and reduces data dimensionality.
SVD
dimensionality
component
Introduction
Python Basics
Basic Mathematical Computations
SVD
Singular value decomposition of a matrix, used to identify principal directions of variation and for compression.
PCA
matrix
compression
Basic Mathematical Computations
Artificial Intelligence: Definition, History, and Future
SVM
Support vector machine, a method that searches for a separating boundary with the maximum margin between classes.
margin
kernel
classification
Prediction Function and Universal Prediction Theory
Perceptron and Support Vector Methods (SVM)