Expresso Churn Prediction Challenge
'%3e%3cg id='b'%3e%3cpath id='a' d='M0-100V0h50z' transform='rotate(18 0 -100)'/%3e%3cuse xlink:href='%23a' transform='scale(-1 1)'/%3e%3c/g%3e%3cuse xlink:href='%23b' transform='rotate(72)'/%3e%3cuse xlink:href='%23b' transform='rotate(144)'/%3e%3cuse xlink:href='%23b' transform='rotate(216)'/%3e%3cuse xlink:href='%23b' transform='rotate(288)'/%3e%3c/g%3e%3c/svg%3e)
Helping Senegal
$1 000 USD
Completed (over 4 years ago)
Classification
Prediction
1378 joined
437 active
Start
Aug 27, 21
Close
Nov 28, 21
Reveal
Nov 28, 21
'/%3e%3cpath fill='%23fff' d='m8.751-17.959 10.11 11.373L3.997-9.844l13.94-6.1-7.692 13.129z'/%3e%3c/svg%3e){kind=small}
Accuracy significantly different in X_test & test
Help · 5 Nov 2021, 08:42 · 4
hey there, i am having a significant difference in accuracy when i test my model on X_test (30% of the train.csv that i am using for accuracy prediction) and test.csv. Like when i used X_test to predict the model's accuracy, its giving of total 0.86 and when i use the same model to predict churn for test.csv and upload the solution, its giving an accuracy of 0.5. Is it normal or its an error indication cause i have cross checked multiple times?
Hello,
The LB evaluation metric is AUC. Could you check the AUC of your local validation set? Note that it's possible to have a very high accuracy but low (~0.5) AUC.
Also make sure your predictions are probabilities between 0 and 1, not absolute 0,1.
thankyou so much, i was not using Area Under the Curve as the evaluation matrix and also using churn as absolute 0 & 1. thanks alot.
Hi ensure you are predicting probabilities instead of absolute values. for most sklearn algorithms, switching 'predict(test)' to 'predict_proba(test)[:,1]' should do the trick
thats exactly what solved my problem. thankyou