Mathematics for AI (Dr Tran Anh Tuan)

Wee Title Content

k
1 Introduction to - Overview of AI Applications
Mathematics for AI - Overview of Applied Mathematics
- Introduction to Algebra
- Introduction to Calculus
- Introduction to Statistics and Probabilities
- Introduction to Graph Theory
- Introduction to Information Theory
- Introduction to Optimization

2-3 Linear Algebra - Matrices and Vectors

- Inner Products
- Matrix Multiplication
- Orthogonal and Diagonal Matrix
- Transpose Matrix and Inverse Matrix in
Normal Equation
- Solving Systems of Linear Equations
- L1 norm or Manhattan Norm
- L2 norm or Euclidean Norm
- Regularization in Machine Learning
- Lasso and Ridge
- Feature Extraction and Feature Selection
- Covariance Matrix
- Eigenvalues and Eigenvectors
- Orthogonality and Orthonormal Set
- Span, Vector Spaces, Rank and Basis
- Determinant and Trace
- Principal Component Analysis (PCA)
- Singular Value Decomposition
- Matrix Decomposition or Matrix
Factorization
 - Affine Spaces

4-5 Calculus - Derivative

- Partial derivative
- Second derivative
- Hessian matrix
- Gradient
- Gradient descent
- Critical points
- Stationary points
- Local maximum
- Global minimum
- Saddle points
- Jacobian matrix

6 Statistics and Probabilities - What Is Data?

- Categories Of Data
- What Is Statistics?
- Basic Terminologies In Statistics
- Sampling Techniques
- Types Of Statistics
Descriptive Statistics
- Measures Of Centre
- Measures Of Spread
- Information Gain And Entropy
- Confusion Matrix
Probability
- What Is Probability?
- Terminologies In Probability
- Probability Distribution
- Types Of Probability
- Bayes’ Theorem
- Gaussian Distribution
- Conjugacy and the Exponential Family
Inferential Statistics
 - Point Estimation
- Interval Estimation
- Estimating Level Of Confidence
- Hypothesis Testing

7 Graph Theory - Directed and Undirected Graph

- Path analysis
- Connectivity analysis
- Community analysis:
- Centrality analysis

8 Information Theory - Information

- Self-Information
- Entropy
- Mutual Information (Joint Entropy,
Conditional Entropy )
- Kullback–Leibler Divergence
- Cross Entropy
- Cross Entropy as An Objective Function
of Multi-class Classification

9 Continuous Optimization - Optimization using Gradient Descent

- Constrained Optimization and Lagrange
Multipliers
- Convex Optimization

10 Linear Regression - Problem Formulation

- Parameter Estimation
- Bayesian Linear Regression
- Maximum Likelihood as Orthogonal
Projection

11 Dimensional Reduction - Maximum Variance Perspective

with PCA - Projection Perspective
- Eigenvector Computation and Low-Rank
Approximations
 - PCA in High Dimensions
- Latent Variable Perspective

12 Density Estimation with - Gaussian Mixture Model

Gaussian Mixture Models - Parameter Learning via Maximum
Likelihood
- EM Algorithm
- Latent-Variable Perspective

13 Classification with Support - Separating Hyperplanes

Vector Machine - Primal Support Vector Machine
- Dual Support Vector Machine
- Kernel
- Numerical Solution

14 Outlier Detection - Interquartile Range (IQR)

- Isolation Forest
- Minimum Covariance Determinant
- Local Outlier Factor
- One-Class SVM

15 Recommendation System - Singular value decomposition techniques -

svd
- Hybrid and context aware recommender
systems
- Factorization machines