search

recodehive / Stackoverflow-Analysis

Stack overflow is a professional community for developers. This repo analysis 3 years of developer Survey done by Stackoverflow and do visualization and predict the salary of Data Scientist in future.
https://stackoverflow-analysis.streamlit.app/
MIT License
221 stars 108 forks source link

Model performance under Gradient boosting #43

Closed Pookie-n-Rookie closed 2 months ago

Pookie-n-Rookie commented 3 months ago

Is your feature request related to a problem? Please describe. Checking for the best algorithms among the various bagging and boosting techniques such as gradient boosting,XG boosting etc..

Describe the solution you'd like Gradient boosting trees can be more accurate than random forest because we can train them to correct each others errors.Aim to decrease bias and model will be weighted according to their performance.

Describe alternatives you've considered Alternatively , i would more for XG boost to do the same as gradient boosting

.

What problem is this feature trying to solve? we might check the dataset with boosting techniques to reach a conclusion.

How do we know when the feature is complete? simply we compare for model accuracy than previous random forest model and current gradient boosting model.

Pookie-n-Rookie commented 3 months ago

@sanjay-kv kindly assign me this issue and i would go through the boosting technique for [FEATURE] Add New ML Algorithms #8

sanjay-kv commented 3 months ago

assigned to u @SwarnenduJUIT

MJakash commented 3 months ago

Hey bhaiya ! i am data science and machine learning enthusiast.I have worked on predictive modelling. After seeing this issue i want to work on this issue.If possible then kindly assign this issue to me.

Pranali3103 commented 3 months ago

Hi @sanjay-kv , I would like to work on this issue , please assign me this issue.

github-actions[bot] commented 2 months ago

This issue has been automatically closed because it has been inactive for more than 30 days. If you believe this is still relevant, feel free to reopen it or create a new one. Thank you!