## Homework Assignments

### Hw1: MLE & MAP estimation

Implement perceptron to classify the given set of documents as one of two class +1 or -1.

Data set: download

### Hw2: SVM Implementation

Implement SMO algorithm to classify the given set of documents as one of two classes +1 or -1.

Data set: download

### Hw3: PLSI Implementation

Classify the given documents into K topics.

In other words, the goal is detecting K topics from the input documents.

You can define K by your own. Try different K and choose the one with the best performance.

Data set: download

### Hw4: Derive LDA

- Derive LDA step-by-step.
- Give sufficient explanation fore each step.
- Submit as digital copy ( may be using LATEX or Word document).

Please refer topic model related slides and other online resources (if necessar).