Issue Type

• Existing code improvement

Proposed Changes

I would like to improve train_test_split(dataset, start, end) function existed in learning.py by adding training and testing set in percentage.

Description

When i am working in learning.py, i found that the function train_test_split has three arguments-

• dataset
• start
• end

Here dataset.examples[start:end] is used for testing and rest for training. I want to reframe the function in following way-

train_test_split(dataset, split = testing_split)

Input

• dataset : Our dataset.
• split : Percentage by which it will be splitted in testing set.

Return

• train : (100 - testing_split) * 100 % of dataset.examples.
• val : testing_split * 100 % of dataset.examples.