Calcule a duração dos passeios de bicicleta com um modelo de regressão no BQML avaliações

I'd add a step to compare predictions with real values: SELECT pred.*, src.duration FROM ML.PREDICT(MODEL bike_model.model_bucketized, ( SELECT start_station_name, start_date FROM `bigquery-public-data`.london_bicycles.cycle_hire LIMIT 100) ) as pred, `bigquery-public-data`.london_bicycles.cycle_hire as src WHERE pred.start_station_name = src.start_station_name and pred.start_date = src.start_date

Piotr K. · Revisado há about 5 years

jared j. · Revisado há about 5 years

ALEXANDRE MENEZES S. · Revisado há about 5 years

Teemu K. · Revisado há about 5 years

Very interresting. Unfortunately the time duration (1 hour) is not sufficient. It should be 2 hours.

Florian T. · Revisado há about 5 years

Giovanni B. · Revisado há about 5 years

Jair M. · Revisado há about 5 years

Harshita B. · Revisado há about 5 years

和美 井. · Revisado há about 5 years

ok

Marcos d. · Revisado há about 5 years

Timothy M. · Revisado há about 5 years

Thomas L. · Revisado há about 5 years

Christopher K. · Revisado há about 5 years

Karthik A. · Revisado há about 5 years

Thomas D. · Revisado há about 5 years

Armando F. · Revisado há about 5 years

Anirudh J. · Revisado há about 5 years

Itishree D. · Revisado há about 5 years

Kiran S. · Revisado há about 5 years

Guillaume G. · Revisado há about 5 years

Martino R. · Revisado há about 5 years

Alberto M. · Revisado há about 5 years

Stefan H. · Revisado há about 5 years

Zeyu Z. · Revisado há about 5 years

Zeyu Z. · Revisado há about 5 years