Improving random forest accuracy

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

Improving random forest accuracy

Sehrish agha
Hello,
I have applied random forest, its accuracy is 99 percent while training and cross validation but it reduces to 80 percent with test set. Please tell me how to improve accuracy with test set. 

_______________________________________________
Wekalist mailing list -- [hidden email]
Send posts to: To unsubscribe send an email to [hidden email]
To subscribe, unsubscribe, etc., visit
https://list.waikato.ac.nz/postorius/lists/wekalist.list.waikato.ac.nz
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html
Reply | Threaded
Open this post in threaded view
|

Re: Improving random forest accuracy

Eibe Frank-3
IReducing the classification model's fit to the training data (i.e., building a simpler model that fits this data less closely) *may* help. Also, if you have *prior* knowledge implying that a predictor attribute behaves differently in the test data than in the rest of the data, it might be better to remove it from the data.

Cheers,
Eibe

On Mon, Nov 25, 2019 at 9:10 PM sehrish Agha <[hidden email]> wrote:
Hello,
I have applied random forest, its accuracy is 99 percent while training and cross validation but it reduces to 80 percent with test set. Please tell me how to improve accuracy with test set. 
_______________________________________________
Wekalist mailing list -- [hidden email]
Send posts to: To unsubscribe send an email to [hidden email]
To subscribe, unsubscribe, etc., visit
https://list.waikato.ac.nz/postorius/lists/wekalist.list.waikato.ac.nz
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html

_______________________________________________
Wekalist mailing list -- [hidden email]
Send posts to: To unsubscribe send an email to [hidden email]
To subscribe, unsubscribe, etc., visit
https://list.waikato.ac.nz/postorius/lists/wekalist.list.waikato.ac.nz
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html