In order for intelligent algorithms to learn something from data, the data must be presented in the best possible way. The process of transforming data and preserving only the most relevant, distinctive features is called feature engineering. It is perhaps the most important step in the data science workflow. This webinar aims to provide an overview of the most commonly used methods, the most common pitfalls for different types of data (sensor data, location data...) and problem sets (prediction, profiling...).
This webinar is part of the Mastercourse 'Data Innovation' 2021, and can be followed as a stand-alone session or in combination with other sessions provided in this mastercourse.
- First session: 20 April 13:00 - 15:30
- Second session: 22 April 13:00 - 15:30