-
Notifications
You must be signed in to change notification settings - Fork 2
Assessment of Model OverFitting or UnderFitting #40
Comments
This is my work on this issue . |
@PhaniRohithaKaza as discussed kindly Collab with @diwakarjaiswal880 , thanks |
Dear Diwakar @diwakarjaiswal880 - thanks a lot for your efforts and having done the Level -1 , of this task . As requested by me on the call right now , need a statistical and mathematical explanation of the underlying phenomenon , thanks |
I found this link for understanding Overfitting in mathematical way. |
Thanks
…On Mon, 25 May 2020 at 3:00 PM, Diwakar Jaiswal ***@***.***> wrote:
I found this link for undurstanding Overfitting in mathematical way.
https://www.coursera.org/lecture/analytics-excel/understanding-why-over-fitting-happens-nlhyf
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#40 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AKH3MR7XBZQQDU6EXOSDRE3RTI3ERANCNFSM4MX77GPQ>
.
|
@diwakarjaiswal880 - Diwakar - your own explanation done to me on the call was much better than the explanation done in the link of CourSera |
1/ Read through the Kaggle and other links provided - come up with a bullet-proof strategy to ensure we dont end up -Overfitting or UnderFitting models.
2/ Dataset for practise can be used from KAGGLE or any other source of your choice .
3/ The actual test of - Overfitting or UnderFitting , will be conducted when Users( end users of the product ) end up creating and Fitting models on the DigitalCognition and PyFinTrader platfroms.
We then need a Validation method / class - process / data pipeline , which warns a - end user - stating that their model has been OverFit or UnderFit , and they can modify - such and such - hyper param's or param's and ensure that they dont end up with a OverFit or UnderFit.
https://www.kaggle.com/artgor/how-to-not-overfit
https://www.kaggle.com/c/dont-overfit-ii/discussion/79903
https://www.kaggle.com/c/dont-overfit-ii/data
https://www.kaggle.com/c/overfitting/data
Post your code here in a Jupyter Notebook along with a CSV file with sample data used .
Any doubts or any further clarifications required kindly contact Rohit - WhatsApp Text = +91-9871050873 or LinkedIn Text = https://www.linkedin.com/in/rohitdhankar/
The text was updated successfully, but these errors were encountered: