关于“通过 BigQuery ML 创建机器学习模型：实验室挑战赛”的评价

Krishna S. · 已于 over 1 year前审核

Maximillian Jonathan P. · 已于 over 1 year前审核

David L. · 已于 over 1 year前审核

Ulu Kagan B. · 已于 over 1 year前审核

Rupali P. · 已于 over 1 year前审核

Siddhartha G. · 已于 over 1 year前审核

Thangaraj S. · 已于 over 1 year前审核

Nigel Malino N. · 已于 over 1 year前审核

Sandy W. · 已于 over 1 year前审核

FRANCISCO JAVIER S. · 已于 over 1 year前审核

Narayana V. · 已于 over 1 year前审核

Andre G. · 已于 over 1 year前审核

Edgar D. · 已于 over 1 year前审核

damien c. · 已于 over 1 year前审核

Patricia M. · 已于 over 1 year前审核

good

Samrat N. · 已于 over 1 year前审核

Caio G. · 已于 over 1 year前审核

Patrick J. · 已于 over 1 year前审核

Srijan M. · 已于 over 1 year前审核

Srijan M. · 已于 over 1 year前审核

Christian O. · 已于 over 1 year前审核

Brian shahputera s. · 已于 over 1 year前审核

I had to use my head a little bit and I got a good feel of how I can use sql(BigQuery) to evaluate ML models on a dataset. I see that it can get very complex very fast. Thanks for keeping it simple. The lab has a nice UI and just the right amount of complexitity to test the system.

웅비 최. · 已于 over 1 year前审核

Kalman T. · 已于 over 1 year前审核

I found Task 4 of this lab to be misleading and frustrating. The instructions explicitly state that I need to "predict the average trip duration for all trips from the busiest bike sharing station in 2019 (based on the number of trips per station in 2019)." Based on these instructions, I wrote queries to: Calculate the busiest station in 2019 based on trip count Use that station to make predictions with the subscriber_type model Filter for Single Trip subscribers and 2019 data Calculate the average predicted duration However, the verification system would only accept a query that used a hardcoded station name '21st & Speedway @PCL', which is never mentioned anywhere in the instructions. This is problematic because: The instructions specifically tell us to base our solution on 2019 trip counts There's no indication we should use a pre-defined station The verification system contradicts the written instructions Students waste time trying to debug technically correct solutions A better approach would be to either: a) Update the instructions to specify using '21st & Speedway @PCL' if that's the required solution b) Or modify the verification system to accept solutions that correctly calculate and use the actual busiest station from 2019 data This discrepancy between instructions and verification criteria creates unnecessary confusion and doesn't reflect real-world data analysis practices where we would actually want to calculate the busiest station rather than assuming it.

Anshul K. · 已于 over 1 year前审核