'BigQueryML: Random Forest Classification
Hello stackoverflow community!
I recently found out that Bigquery ML does not support random forest classification models. To overcome that, I figured that I might be able to build a model with sklearn package and then use the same hyperparameters on bigqueryml. Is using the boosted tree model in bigquery the best option in this case?
Solution 1:[1]
You can train a random forest model using BQML's boosted tree model type, with the following parameters
- Set BOOSTER_TYPE to GBTREE
- Set NUM_PARALLEL_TREE to some number greater than 1
- Set MAX_ITERATIONS to 1
- Set LEARN_RATE to 1
- Set one of {COLSAMPLE_BYTREE, COLSAMPLE_BYLEVEL, COLSAMPLE_BYNODE} less than 1.
Sources
This article follows the attribution requirements of Stack Overflow and is licensed under CC BY-SA 3.0.
Source: Stack Overflow
| Solution | Source |
|---|---|
| Solution 1 | Xi Cheng |