Fixed #2, train and test a classification model on vehicles dataset

[Bolaji61]

No description provided.

[dzeber]

This is a great start on this task. I am happy to see the diversity of models you are looking into.

I have a couple of recommendations:

• Please remove .DS_Store from your PR. That is supposed to be excluded automatically by the .gitignore, but I just noticed it's not included in there - I will fix the gitignore on the master branch.
• The code you use to test out the models should be moved to a separate module and called from there (and that way you can also reuse the same code as a single function).
• I saw you included comments in the notebook explaining what your code does. That is good - however, I'd also like to see more text/markdown cells explaining your thought process: why you decided to do certain analyses, and your interpretations of the results. This is a great help to someone reading your notebook for the first time. For example, the correlation plot shows some fascinating patterns. It would be great to include a few sentences discussing your conclusions from seeing this - what should we pay attention to and how did it influence the rest of your work.
• The accuracy of 24% for the SVM looks suspect - I'd double-check to make sure it is working correctly. That's no better than guessing class labels at random.

[Bolaji61]

Train and Test a Classification Model on vehicles.csv dataset.
I've been able to create a separate modules file containing the models and functions used.
Next is to study the performance evaluation metrics and give detailed explanations of each line of code.
@dzeber , I hope my pace is not too slow, I'm quite new to open source. Thanks for your earlier review.

[Bolaji61]

I have properly documented my thought process on my choice of model and test_size fraction. Next, is to tune the parameters of the models chosen in order to achieve a higher accuracy.

[dzeber]

Thanks for these updates - this has developed into a nicely presented, well-documented notebook. I'm merging this now, as it satisfies the requirements for #2. If you wish to continue working on it, I recommend checking out the other issues in the repo.

A couple of general comments (don't need to fix at present):

• The _model() functions in your modules.py have a lot of repeated code. An improvement would be to combine into a single function and supply the parts that change as params.
• For the presentation of your very nice analysis of test split sizes, rather than hard-coding the results to present in the table, why not run the evaluations in the notebook, capture them dynamically and present the results directly.

[Bolaji61]

Yaay, this is great news. Thanks a lot, its my first ever merged PR in an Open Source project. I'll work on other issues in the project and fix the issues you commented on later as suggested.
I do appreciate your reviews a lot, I hope to help some others this same way when I'm advanced. Do have a great weekend.

[Bolaji61]

Thanks for these updates - this has developed into a nicely presented, well-documented notebook. I'm merging this now, as it satisfies the requirements for #2. If you wish to continue working on it, I recommend checking out the other issues in the repo.

A couple of general comments (don't need to fix at present):

* The `_model()` functions in your modules.py have a lot of repeated code. An improvement would be to combine into a single function and supply the parts that change as params.

* For the presentation of your very nice analysis of test split sizes, rather than hard-coding the results to present in the table, why not run the evaluations in the notebook, capture them dynamically and present the results directly.

The first suggestion has been fixed accordingly