Bug Review.

[binguliki]

@Ramani25 and @aksha98, I can see that in the recent changes you have made for model selection and testing, there are some bugs:

1) Data Splitting Before Sampling:

Here, you people have performed data splitting before sampling. Like, what's the point of sampling then? Now because of this, there is a great imbalance in the data, which leads to poor model performance.

---

2) Scaling Data Twice in SVC Model:

When you trained the SVC model, why did you scale the data again? I understand that you have used ChatGPT to train the model, but at least don't copy-paste the code entirely. Try to understand what's happening at each step.

---

3) Saving Preprocessing Pipeline and Model:

After all of the bugs have been resolved, please save the preprocessing pipeline and the model. I have created an issue #9 regarding this. So please link it with the next pull request that you make.