search

prateekiiest / titanic_survival_exploration

Udacity Machine Learning Nano degree Program Project Predicting Passenger Survival
MIT License
21 stars 35 forks source link

Making compound Features out of input Features #18

Closed vishal1383 closed 6 years ago

vishal1383 commented 6 years ago

Thank you for your contribution. Please provide the details requested below.

ISSUE NUMBER

https://github.com/prateekiiest/titanic_survival_exploration/issues/9

SHORT DESCRIPTION

I have made a compound feature inspired from Kaggle contestants Sinai,Anisotropic who also make use of the Name feature from which they extract out title using Regular Expression matching.Title also contains the Information of Sex and some more extra information also.

TESTING

Can run the code on a jupyter notebook.

vishal1383 commented 6 years ago

Will check and fix that as well Thanks 1.Actually, I added the code to obtain the histogram and forgot to delete after that .Thanks for pointing out. 2.I added the column Title in the data frame in the input on my local machine.Will modify it in the code to add a new column and correlation 3.My bad will fix that too. Thanks !! for pointing out.

vishal1383 commented 6 years ago

This shouldn't happen .Checking...

vishal1383 commented 6 years ago

Have to make those changes again I guess..

vishal1383 commented 6 years ago

Thanks, will add the Pearson Correlation matrix and the accuracy obtained using the Title Feature also.

prateekiiest commented 6 years ago

And @imVivekGupta , thanks for the helping hand :)