ENGR 421 • Final • Introduction to Machine Learning

Counting and Probability

Conditional Probability and Independence

Bayes' Rule

Discrete Random Variables

Continuous Random Variables

Expected Value and Variance

Bernoulli and Binomial Distributions

Continuous Uniform Distribution

Exponential Distribution

Normal Distribution

Laplace and Logistic Distributions

Dimensionality Reduction

Principal Component Analysis (PCA)

PCA: Choose k Using Proportion of Variance

Feature Embedding & Factor Analysis (FA)

Singular Value Decomposition and Matrix Factorization

Multidimensional Scaling

Linear Discriminant Analysis (LDA)

Canonical Correlation Analysis

Isomap, Locally Linear Embedding, Laplacian Eigenmaps

Which Dimensionality Reduction Method and Why?

Introduction and Mixture Densities

K-Means Clustering

One Iteration of k-Means

Expectation-Maximization (EM)

Mixture Models & Practical Use of Clusters

Spectral and Hierarchical Clustering

Choose the Right Clustering Tool

“Clustering as Preprocessing” Pitfall

Why Combine Multiple Learners?

Voting & Linear Combination

Bayesian Perspective & Effect of Dependence

Fixed Combination Rules & ECOC

Bagging & AdaBoost

Mixture of Experts and Stacking

Fine-Tuning an Ensemble

Cascading

Combining Multiple Sources/Views

Which Ensemble Method Fits?

Correlation vs Ensemble Gain

Why do we run ML experiments?

Algorithm preference

Factors & Response

Guideline

Spot the Leakage: Is This Cross-Validation Setup Valid?

Fix the Experiment: Where Does Each Step Belong?

Pass Rates & Majors (Bayes; Law of Total Probability)

Weighted Least Squares (Closed-Form Solution, Matrix View & Interpretation)

MLE for α (positive support, exponential tail)

Linear Discriminant with Equal Variance

MLP with Hard-Threshold Units

Should we initialize all MLP weights to zero?

One Shared Network vs. Three Separate Networks

Naive Histogram Estimator vs. Parzen Windows (Kernel)

Kernel Smoother

Naive Density Estimator (Bandwidth effect & validity)

Comparing Two Splits (Gini vs. Misclassification)

Prepruning vs. Postpruning (Which and Why?)

Why Not Regression for Classification?

From Binary to Multiclass: One-vs-All / One-vs-One with a Binary Classifier

Max-shift for SoftMax

Why Initialize Weights Near Zero?

Adaptive Learning Rates in Gradient Descent

When Do Direct Input Output Links Help in an MLP?

Mahalanobis vs. Euclidean: Why and When?

Discrete Attribute in Decision Trees

Regularized Least Squares

Gaussian Generative Model → Logistic Posterior

Naive Bayes Text Classification with Binary Features

Derivative of Softmax

Kernel Density Estimation

Choosing Between Two Splits: Gini vs. Misclassification

Mean Square Error for Linear Regression

Gradient Descent Update

k-NN Regression Prediction

Decision Boundary and Building a Network for Binary Classification

Derivative of Squared Error

Computing Input and Output of a Convolution Node

True/False Reasoning on Activation, Linear Networks, and Gradient Descent

Computing Total Probability

True/False on Scaling, k-NN, Intrinsic Error and Model Complexity

Output Size of a Conv Layer

Regression: Test-Set MSE

Generalization & Overfitting: True/False

Baseline Error: ZeroR vs Random Guessing

Entropy: Fair Die & Bias Effect

Decision Trees: ID3 Optimality + Key Advantage

Decision Tree Split: Remaining Entropy