CS 189 Review

CS 189 Review#

Probabilistic Models#

Discriminant Analyisis#

Regression#

Decision Trees#

  • Object, Algorithm

  • Random Forest

  • Adaboost

Kernelization#

  • Ridge Kernel

  • Polynomial Kernel

Neural Networks#

  • Fully Connected Layer

  • Convolution Layer

Clustering#

  • k-Means Clustering

  • Hierarchical Clustering

  • Spectral Clustering

k-Nearest Neighbor#

Take the goal, algorithmm and use case of:

  • Vornoi Diagram

  • k-d Tree

  • Exhausive Search

Principle Component Analysis#

  • Latent Factor Analysis

Singular Value Decomposition#

  • Definition

  • Low Rank Approximation

Feature Engineering#

  • Feature Selection

  • Dimensionality Reduction

  • Feature Scaling

Page 1#

Heuristics#

  • Neural Network

    • \(\downarrow\) Training Error: layers, nodes/units, normalization

    • CNN Intuition: Last layers are components of an image

    • Normalization: LASSO, neural network, SVM

Optimization#

  • Gradient Descent

  • Convexity

  • Loss Functions

    • Cross Entropy

  • Line Search Methods

    • Newton-Raphson’s Method

  • Quadratic Programs

Linear Algebra#

  • Calculus

Contents