Deploy And Train Azure Machine Learning Model

This article continues from the last article about How to create Azure Machine Learning Datasets.

After that, create the Azure Machine Learning Model Deploy and Train. Cloud computing's scalability allows Azure Machine Learning to run several preprocessing techniques and model-training algorithms in parallel to seek out the simplest performing supervised machine learning model for your data.

Build an automated machine learning experiment

In Azure Machine Learning, operations that you simply run are called experiments.

Follow the steps below,

Steps:

In the Datasets section,

Select the generated Datasets file option.

Then enter into Configure run:

New experiment name: Created file.

Target column: rentals (this is that the label the model are going to be trained to predict)

Select compute cluster: the compute cluster you created previously

After that, Select task and settings:

Task type: Regression (the model will predict a numeric value)

Then, under task type, there are settings View additional configuration settings and View Featurization settings. need to configure these settings next.

Select and open the Additional configuration settings:

Primary metric: Select Normalized root mean squared error (more about this metric later!)

Explain the best model: Selected - this feature causes automated machine learning to calculate feature importance for the simplest model; making it possible to work out the influence of every feature on the anticipated label.

Blocked algorithms: Block all aside from RandomForest and LightGBM - normally you'd want to undertake as many as possible but doing so can take an extended time!

Exit criterion:

Training job time (hours): 0.5 - this causes the experiment to finish after a maximum of half-hour.

Metric score threshold: 0.08 - this causes the experiment to finish if a model achieves a normalized root mean squared error metric score of 0.08 or less.