Creating Machine Learning Frameworks

This month we are delighted to welcome Dr Andriy Temko to present at Cork AI. Andriy will be presenting a tutorial on how to set up machine learning frameworks for experiments. The tutorial will cover multiple topics, including rigor on handling data, what should be used for train, validation and test, and what questions can be answered by each data stream.

Establishing a good comparative framework will help to distinguish result improvements which happen due to genuine technical contributions from those which happen due to data manipulation, metrics changes, overfitting, etc. It also help to better track progress in a quantitative manner.

Andriy is a principal engineer and machine-learning lead in Qualcomm in Cork, working on the application of machine learning to System-on-Chip design processes. Andriy has a PhD Degree in Telecommunication from Technical University of Catalonia, and also holds an adjunct senior lecturer position in School of Engineer in UCC. Prior to joining Qualcomm, he worked for 10 years in the INFANT research centre on development of AI diagnostic solutions for neonatal healthcare.