How to use parameter weight_column_name in AutoMLConfig Class for Automated Azure MAchine Learning

Question

How to use parameter weight_column_name in AutoMLConfig Class for Automated Azure MAchine Learning

Alexandros Papaspyridis 1

Hi everybody,
could you please help with the weight_column_Name in the AutomLconfig how is used,and how should the syntax be for multiple columns? We are trying to run various different experiments with specific weights in some columns that we consider more serious. Can somebody provide an example of this parameter in use?

Kind Regards

romungi-MSFT 48,906 Reputation points Microsoft Employee Moderator

2022-01-18T14:37:02.037+00:00
@Alexandros Papaspyridis I think multiple column names cannot be included in this case. As per the reference description it only takes the column name of your training or validation data.

The name of the sample weight column. AutoML supports a weighted column as an input, causing rows in the data to be weighted up or down. If the input data doesn't have column names, column indices can be used instead, expressed as integers.

example:

automl_settings = { "primary_metric": AUTOML_METRIC, "enable_early_stopping": True, "experiment_timeout_hours": TIMEOUT, "max_concurrent_iterations": 4, "max_cores_per_iteration": -1, "verbosity": logging.INFO, "enable_voting_ensemble": False, "enable_stack_ensemble": False, "weight_column_name": "weight_col_name" }
Alexandros Papaspyridis 1 Reputation point

2022-01-19T14:02:13.693+00:00

thanks a lot @romungi-MSFT for the clarification. And may i ask what weight does it enlist for the column in the parameter?
Thanks a lot

2 answers

Your answer

romungi-MSFT 48,906 Reputation points Microsoft Employee Moderator

2022-01-18T14:37:02.037+00:00

@Alexandros Papaspyridis I think multiple column names cannot be included in this case. As per the reference description it only takes the column name of your training or validation data.

The name of the sample weight column. AutoML supports a weighted column as an input, causing rows in the data to be weighted up or down. If the input data doesn't have column names, column indices can be used instead, expressed as integers.

example:

automl_settings = { "primary_metric": AUTOML_METRIC, "enable_early_stopping": True, "experiment_timeout_hours": TIMEOUT, "max_concurrent_iterations": 4, "max_cores_per_iteration": -1, "verbosity": logging.INFO, "enable_voting_ensemble": False, "enable_stack_ensemble": False, "weight_column_name": "weight_col_name" }
Alexandros Papaspyridis 1 Reputation point

2022-01-19T14:02:13.693+00:00

thanks a lot @romungi-MSFT for the clarification. And may i ask what weight does it enlist for the column in the parameter?
Thanks a lot

Answer 1

Kaushik Jeyaraman 1

how do I add this setting with the autoML UI?

Answer 2

Kaushik Jeyaraman 1

how do I set this in the auto ML UI?

Share via

How to use parameter weight_column_name in AutoMLConfig Class for Automated Azure MAchine Learning

2 answers

Your answer