Data Augmentation plays a prominent role in improving the generalization performance of machine learning solutions. This talk will first introduce the scientific rationale behind data augmentation and then will delve into the best practices of its utilization with real-life examples in Python. Specific data augmentation techniques for computer vision, natural language processing, and time series problems will be discussed with relevant open-source libraries. Finally, test-time data augmentation and data augmentation in the context of algorithmic fairness will be covered.

Oguzhan Gencoglu is the Co-founder and Head of AI at Top Data Science, a Helsinki-based AI startup that provides AI development as a service. With his team, he delivered more than 70 machine learning solutions in numerous industries for the past 4.5 years. Before that, he used to conduct machine learning research in several countries including USA, Czech Republic, Turkey, Denmark, and Finland.


Oguzhan Gencoglu - Best Practices for Data Augmentation
1.55 GEEK