Solved–Homework 6 –Solution

Description

Note

HW6 is optional. We will select your highest 5 homework grades to calculate your nal homework grade (up to 25% based on the grading policy).

You are expected to submit both a report and code. The submission format is speci ed on CCLE under HW6 description.

Copying and sharing of homework are NOT allowed. But you can discuss general challenges and ideas with others. Suspicious cases will be reported.

\# ======== YOUR CODE HERE =========” is used where input from you is needed in the code le.

In Homework 6, you will complete two basic text mining model in lectures: Naive Bayes and Topic modeling (pLSA). Python 3.6 is recommended in HW6 and you may need to install additional packages such as sklearn and jieba (numpy, pandas and matplotlib are also needed.) Besides, you are allowed to slightly change the code or add new functions other than \YOUR CODE HERE” as long as you keep the structure and do not use external model toolkits.

You need to submit all your python les, report and output les (including gures and text les). Do NOT include original dataset les.

• Naive Bayes for Text (50 points)

Naive Bayers is one generative model for text classi cation. In the problem, you are given a document in dataset folder. The original data comes from \20 newsgroups”. You can use the provided data les to save e orts on preprocessing.

1. Complete the implementation of Naive Bayes model for text classi cation in nbm.py. After that, run nbm sklearn.py, which uses sklearn to implement naive bayes model for text classi cation (Note that the dataset is slightly di erent).

2. Report your classi cation accuracy on train and test documents. Also report your classi cation matrix. Show one example document that Naive Bayes classi es incorrectly ((you can ll in the following table).

3. Question: Is Naive Bayes a generative model or discriminative model and Why? What is the di erence between Naive Bayes classi er and Logistic Regression? What are the pros and cons of Naive Bayes for text classi cation task?

Introduction to Data Mining (UCLA CS 145) Homework 6

Table 1: Report accuracy for Naive Bayes Model

Train set accuracy Test set accuracy

sklearn implementation

your implementation

Table 2: Incorrect Examples

Words (count) in the example document	Predicted label	Truth label
For example, student(4), education(2), etc	Class A	Class B

4. Question: Can you apply Naive Bayes model to identify spam emails from normal ones? Brie y explain your method.

• Topic Modeling: Probabilistic Latent Semantic Analysis (pLSA) (50 points)

In this section, you will implement Probabilistic Latent Semantic Analysis (pLSA) by EM algorithm.

1. Complete the implementation of pLSA in plsa.py. You need to nish the E step, M step and likelihood function.

2. Choose di erent K (number of topics) in plsa.py. What is your option for a reasonable K in dataset1.txt and dataset2.txt? Give your results of 10 words under each topic by lling in the following table (suppose you set K = 4).

Table 3: Topic words

Dataset 1

Topic 1 Topic 2 Topic 3 Topic 4

Dataset 2

Topic 1 Topic 2 Topic 3 Topic 4

3. Question: Are there any similarities between pLSA and GMM model? Brie y explain your thoughts.

3. Question: What are the disadvantages of pLSA? Consider its generalizing ability to new unseen document and its parameter complexity, etc.

2