Agent Platform: Predicting Loan Risk with AutoML Reviews

Kuo-Tai T. · Reviewed almost 2 years ago

Cesar Enrique V. · Reviewed almost 2 years ago

Janet W. · Reviewed almost 2 years ago

Gavriel D. · Reviewed almost 2 years ago

Nice!

Diego E. · Reviewed almost 2 years ago

Harsha B. · Reviewed almost 2 years ago

Camila M. · Reviewed almost 2 years ago

Jorge R. · Reviewed almost 2 years ago

Jerry X. · Reviewed almost 2 years ago

Model training takes too long, giving not enough time to complete the session

Yuri D. · Reviewed almost 2 years ago

Julian B. · Reviewed almost 2 years ago

Scott T. · Reviewed almost 2 years ago

Giulio S. · Reviewed almost 2 years ago

Melissa R. · Reviewed almost 2 years ago

Prateek C. · Reviewed almost 2 years ago

nice

Yell Yint M. · Reviewed almost 2 years ago

Syam G. · Reviewed almost 2 years ago

Deskitex G. · Reviewed almost 2 years ago

Ferry Yun K. · Reviewed almost 2 years ago

No matter what combinations of test data I attempted, the model always responded with a model class = 0. It did change the model scores quite a bit. However, in the explanations, an example gives a model_class=0 with a model_score in the order of 1E-5 and states that the model predicts the user will pay the loan, which doesn't make a lot of sense.

Alessandro T. · Reviewed almost 2 years ago

Tom H. · Reviewed almost 2 years ago

Cosmin A. · Reviewed almost 2 years ago

Difa S. · Reviewed almost 2 years ago

Why not train and use the actual model. The approach used is not very enlightning.

J. S. · Reviewed almost 2 years ago

Ahmad G. · Reviewed almost 2 years ago