Dados de engenharia para modelagem preditiva com o BigQuery ML: laboratório com desafio avaliações

Abhijit p. · Revisado há over 1 year

Kevin C. · Revisado há over 1 year

Ruturaj D. · Revisado há over 1 year

Nico J. · Revisado há over 1 year

Ishwari M. · Revisado há over 1 year

Amar K. · Revisado há over 1 year

I Kadek Virgio Adrian K. · Revisado há over 1 year

Atul K. · Revisado há over 1 year

Anurag Singh R. · Revisado há over 1 year

anandh k. · Revisado há over 1 year

Rekha T. · Revisado há over 1 year

Saranya M. · Revisado há over 1 year

Abidi R. · Revisado há over 1 year

Onio G. · Revisado há over 1 year

Rahmad N. · Revisado há over 1 year

Kanad S. · Revisado há over 1 year

Rajjo G. · Revisado há over 1 year

Cecilia Z. · Revisado há over 1 year

The first time I did everything right, but didn't have enough time to do the last task. The second time, I did the same as I did before, but couldn't get the RMSE down below 10 even though I used the same exact query to create the model. This seems to be because the parameters to filter on changed so there was too little data to train on. I upsampled to as close to 1 million as I could and still nothing would work for linear regression. I even tried automl_regression, but the model takes hours to train and we only have 1 hour. The third time I did it I was able to complete the entire thing in <15 minutes with everything the same as the first time because the filter parameters allowed more data through. Please fix this so all generated filter parameters should allow the user to create a model that gets an RMSE below 10.

Ben S. · Revisado há over 1 year

Habib A. · Revisado há over 1 year

Aakif M. · Revisado há over 1 year

Rishabh S. · Revisado há over 1 year

Govind S. · Revisado há over 1 year

Govind S. · Revisado há over 1 year