Kaggle Problem: Titanic - Basic ML Implementation

[vibhorkrishna]

I have done this problem using various ML models and have used only basic knowledge of ML so that it is easy for beginners to understand. I have also explained the problem in readme file. Please add hacktoberfest-accepted label to this pull request

[abhinand5]

Thanks for your contribution @vibhorkrishna

I've reviewed your work, it looks good, but it would be even better if you can make these few changes...

1. Can you bring all that into a single jupyter notebook? (You can still keep the scripts if you want)
2. Can you write sections of your code in functions to avoid repeating the same chunk of code for multiple algorithms?
3. Optional - EDA with output graphs

[vibhorkrishna]

I've made the following changes:

1. Brought all code into a single jupyter notebook
2. Written the code for multiple algorithms in functions
3. Placed some graphs. Was unable to place output graphs as I don't know the correct output of test dataset. Therefore was unable to make confusion matrix.

[abhinand5]

Great work @vibhorkrishna Once you do the following changes I'll merge it, thanks a lot for your contribution!

• [ ] You can just pull out y_test from test dataset as you have done for X_test and then you can use sklearn metrics module for (y_test, y_pred) to compute accuracy, confusion matrix and whatever you want.
• [ ] Do all the import statements at the start of the notebook.

[vibhorkrishna]

There is no y_test in test dataset. We only know the accuracy of the y_pred when we submit y_pred to kaggle submissions. Otherwise I would have calculated confusion matrix. I have written the accuracy of each model in README file by submitting my y_pred values one by one for each model in Kaggle.

[abhinand5]

Got it. Then,

• [ ] train_test_split from the train data itself and use that for validation.
• [ ] Do all the import statements at the start of the notebook.

[vibhorkrishna]

I've made the following changes:

1. Split the data using train_test split, trained the data and calculated the accuracy and confusion matrix
2. Made all the import statements at the start of the notebook

[vibhorkrishna]

As this issue is not in hacktoberfest. Therefore please label my pull request as hacktoberfest-accepted

[abhinand5]

Sure, Thanks for your contribution @vibhorkrishna

[abhinand5]

@vibhorkrishna Leave a star to help out this repo's visibility, if interested.

[vibhorkrishna]

Done