Tutorial: IBM Watson Studio AutoAI: Modeling for the rest of us

AutoAI makes machine learning and AI accessible to all. Let's explore what is required to use it and create models. Then we can see how we can use the model through Watson Studio or programmatically through a notebook using Python.

Tutorial

Creating models manually is difficult. It starts with finding candidate algorithms that best fit the specific case. Do we know all the appropriate algorithms? Then we have to prepare the data by converting any non-numeric fields to numerical values. Do we need to do additional feature engineering? How do we tune all the hyper-parameters for each chosen algorithms?

AutoAI takes care of all those steps for us and gives us a set of ranked models to choose from. With all the details provided, choosing the best model becomes easy and we can go on with implementing our business solution.

In this tutorial, you will explore the following key capabilities:

• Many algorithms may fit the business problem
• Model creation may require additional feature engineering
• Model creation requires hyper-parameters tuning
• AutoAI removes the complexity of model creation