Setting up Automated ML Training

This guide offers a comprehensive step-by-step approach to setting up automated machine learning training, making it easier for users to implement ML solutions efficiently.

Hafeez Baig
16 steps
2 minutes

Azure

This lab is a continuation of the previous one, Please refer to it before proceeding - <https://scribehow.com/viewer/Automated_Machine_Learning__Dataset_and_Compute__7i9CwjoOTAeJLe4DjsbWAA>

On the **Azure AI** page, click on the **Automated ML** option from the left bar

**What is Automated ML in Azure AI?** **Automated ML (AutoML)** in Azure AI is a tool that automates the process of selecting, training, and tuning machine learning models. Users simply provide a dataset and define the target column, and AutoML handles preprocessing, algorithm selection, and hyperparameter tuning. It’s ideal for users with minimal ML experience, enabling faster and more accurate model development.

**Automated ML** page will open, click on the **New Automated ML job** button on the top left side

**Training method** tab will open, **Job name** will be auto generated

Scroll the page and click on **Next** button

**Task type & data** tab will open, click and select **Regression** from the **Select task type** dropdown list

**What is Regression?** **Regression** is a type of supervised machine learning used to predict continuous numeric values based on input features. It models the relationship between variables, for example, predicting house prices based on size and location, or forecasting sales over time. Common regression algorithms include Linear Regression, Decision Trees, and Gradient Boosting.

In the **Select data** section select the **my-dataset-1** which we have created earlier and click on the **Next** button

Task settings tab will open, Select the **Target column** as **Y (Integer)**

**What is Y (Integer)?** **Y (Integer)** typically refers to the **target variable (label)** in a machine learning regression or classification task, where the output values are integers. - In **classification**, Y could represent categories encoded as integers (e.g., 0 = cat, 1 = dog). - In **regression**, it might be a count or discrete numeric output (e.g., number of purchases). It defines what the model is trying to predict.

Click on the **View additional configurations** **settings** link below the **Target column** **Note:** Blocked models in Azure AI refer to models that are restricted from use within an Azure environment

**Additional configurations** wizard will open on the right side, click on the **Blocked models** dropdown

Select all the options in the blocked models dropdown except **ElasticNet, GradientBoosting, DecisionTree** and **KNN,** and click on the **Save** button

In the **Task settings** tab, click on **View featurization settings** link under **Target column** section

**Featurization** wizard will open, here you can modify the featurization

**What is Featurization?** **Featurization** is the process of transforming raw data into meaningful input features that a machine learning model can understand and learn from. It includes tasks like: - Converting text to numerical values (e.g., using embeddings or token counts) - Encoding categorical variables (e.g., one-hot encoding) - Normalizing or scaling numeric data - Extracting date/time components (e.g., day, month) Effective featurization improves model performance by highlighting important patterns in the data.

Scroll the page and click on **Next** button

**Compute** tab will open, keep everything as default and click on the **Next** button

Review tab will open, review the settings and click on the **Submit training job** button

**Job Overview** page will open, here you can see the details **Note:** Creation of the Job will take sometime

Congratulations on completing this lab and setting up Automated ML Training!

Refer to this lab to continue learning - <https://scribehow.com/viewer/Exploring_Automated_ML_Training__3EwGJMLMTYSJ4hVvwnZMsQ>