Information
Section: Biased data
Goal: Understand how biased data affect the prediction and how to fix it.
Time needed: 80 min
Prerequisites: AIS data, basics about machine learning
Biased dataΒΆ
In this section, we will explore how the data can be misused to solve prediction problems, and how to use them in the best possible way.
We will go through with three parts:
Type of task: how to choose between regression and classification, depending on the attribute we want to predict and its meaning in the real world.
Attributes selection: how to carefully select the right attributes for solving the problem.
Prediction from splitted dataset: how, sometimes, the available data can be splitted in different parts to build different and more performant models.