rohitinu6 / Stock-Price-Prediction

This project focuses on predicting the stock prices of "The State Bank Of India" using machine learning Regression algorithms.
MIT License
46 stars 80 forks source link

Code Optimization Proposal - Repeated Scaling and Splitting #21

Closed AYUSHI-SHA closed 1 month ago

AYUSHI-SHA commented 1 month ago

Hi, I am Ayushi Sha. I’ve been reviewing the project as part of my contribution to GSSoC Extended 24 and noticed that the code performs scaling and splitting multiple times. To optimize the workflow and improve efficiency, I’d like to propose making changes that involve scaling and splitting the data just once, and reusing the processed data afterward.

This would help reduce redundancy and potentially enhance performance. I would like to work on this. If you like this idea, please assign the task to me and add the corresponding labels to it. Thank you for your time and consideration!

Screenshot 2024-10-03 163110

image

image

AYUSHI-SHA commented 1 month ago

@rohitinu6 I noticed that feature scaling has been performed on the data set but to make predictions instead of using X_test_scaled everywhere, X_test is being used which is significantly changing the results. I think it's a typography mistake and for now I am changing it to X_test_scaled. Do get back to me if it is to be left to X_test only.