Local Model-Agnostic Explanations for Machine Learning and Time-series Forecasting Models

Abstract

As we rely more and more on Artificial Intelligence machine learning models for real-life decision-making, understanding and trusting the predictions becomes ever more critical. Imagine that when you buy your first home, the bank rejects your mortgage application without any reason. It is practically unethical to make decisions based on automated systems without providing explanations. This refers, how crucial it is to know how and why the system arrived at a conclusion. This is where my Ph.D. comes into play. In my Ph.D. I developed novel algorithms which provide human-understandable explanations for the decisions of classification and forecasting models.