Machine Learning Foundations/Techniques (NTU, Fall 2020)

• Course website
  • https://www.csie.ntu.edu.tw/~htlin/course/ml20fall/
• Abstract of HWs
  • hw1
    • The Learning Problem
    • Perceptron Learning Algorithm (PLA)
    • Off-Training-Set Error
    • Hoeffding Inequality
    • Bad Data
    • Multiple-Bin Sampling
    • Experiments with Perceptron Learning Algorithm
  • hw2
    • Perceptrons
    • Ring Hypothesis Set
    • Deviation from Optimal Hypothesis
    • The VC Dimension
    • Noise and Error
    • Decision Stump
  • hw3
    • Linear Regression
    • Likelihood and Maximum Likelihood
    • Gradient and Stochastic Gradient Descent
    • Hessian and Newton Method
    • Multinomial Logistic Regression
    • Nonlinear Transformation
    • Experiments with Linear and Nonlinear Models
  • hw4
    • Deterministic Noise
    • Learning Curve
    • Noisy Virtual Examples
    • Regularization
    • Leave-one-out
    • Learning Principles
    • Experiments with Regularized Logistic Regression
  • hw5
    • Hard-Margin SVM and Large Margin
    • Dual Problem of Quadratic Programming
    • Properties of Kernels
    • Kernel Perceptron Learning Algorithm
    • Soft-Margin SVM
    • Experiments with Soft-Margin SVM
  • hw6
    • Neural Networks
    • Matrix Factorization
    • Aggregation
    • Adaptive Boosting
    • Decision Tree
    • Experiments with Decision Tree and Random Forest