Q$1.$ Naive Bayes: Code $[25]$

In this question, you will learn to build a Naive Bayes Classifier for the binary classification

task.

Dataset: "Financial Phrasebank" dataset from HuggingFace. ${?}^1$ To load the data, you

need to install library "datasets" $($pip install datasets$)$ and then use load$\_$datset $()$

method to load the dataset. You can find the code on the link provided above.

The dataset contains $3$ class labels, neutral $(1),$ positive $(2),$ and negative $(0).$ Consider

only positive and negative samples and ignore the neutral samples. Use $80\%$ of the

samples selected randomly to train the model and the remaining $20\%$ for the test.

Clean the dataset with the steps from the previous assignment and build a vocabulary of

all the words.

Compute the prior probability of each class

$p(c_i)=\frac{\text{c}\text{o}\text{u}\text{n}\text{t}(c_i)}{N}$

Here, count$(c_i)$ is the number of samples with class $c_i$ and $N$ is the total number of

samples in the dataset.

Compute the likelihood $p(w_i|c)$ for a all words $w_i$ and all classes $c$ with following equation:

$p(w_i|c)=\frac{\text{c}\text{o}\text{u}\text{n}\text{t}(w_i,c)+1}{|V|+{\displaystyle \underset{w_{}V}{\overset{?}{}}}\text{c}\text{o}\text{u}\text{n}\text{t}(w,c)}$

Here, the count$(w_i,c)$ is the frequency of the word $w_i$ in class $c$ while ${\displaystyle \underset{wV}{\overset{?}{}}}$count$(w,c)$ is

the frequency of all the words in the class $c.$ Laplace smoothing is used to avoid zero

probability in the case of a new word.

For each sample in the test set, predict class $c_{NB}$ which is the class with the highest

posterior probability. To avoid underflow and increase speed, use log space to predict

the class as follows:

$c_{NB}=argma{x}_{cC}(logp(c)+{\displaystyle \underset{w_{i}V}{\overset{?}{}}}logp(w_i|c))$

Using the metrics from scikit$-$learn library ${?}^2,$ calculate the accuracy and macro$-$average

precision, recall, and F$1$ score, and also provide the confusion matrix on the test set.