README

hw3

Principal Component Analysis and N-gram Language Models

Overview

1. Packages Needed for this Project

2. Dataset

3. Program Specification

4. Part A: Principal Component Analysis ([60] points)

4.1 Load and Center the Dataset ([8] points)

4.2 Find the Covariance Matrix ([8] points)

4.3 Get the m Largest Eigenvalues and their Eigenvectors ([15] points)

4.4 Get all Eigenvalues/Eigenvectors that Explain More than a Certain proportion of the Variance ([10] points)

4.5 Interlude: Understanding PCA

4.5.1 PCA Projection

4.5.2 PCA Reconstruction

4.6 Project the Images ([9] points)

4.7 Visualize ([10] points)

5. Part B: N-gram Language Models ([40] points)

5.1 Background: N-gram Language Models

5.2 Functions to Implement

5.3 Example Usage

5.4 Testing Your Implementation

6. Deliverables