AIAI Seminar - 21/06/21 - Tiejun Ma

Speaker: Tiejun Ma  

Title: Human, Technologies and AI enhanced Decision Making

Abstract: Artificial Intelligence aims to “simulate human intelligence in machines, allowing computer programmes to think like humans and make decisions”. However, scientific communities still have limited understanding on how humans behave and make decisions.  Part of such a challenge is associated with having effective ways of measuring the quality of human decisions and understanding human’s risk-taking behaviour when facing real-world uncertainties. Quantitative measuring the performance of a human’s decision is a challenging task, the availability of suitable data and methodologies are major issues. In this introductory talk, I will provide a high-level overview of some of my group’s previous and recent research work related to understanding how human’s decisions are impacted by technologies. I will briefly overview three pieces of case studies of my group’s research work and financial transaction data that are adopted as an observation window of human’s decisions.  First, my research group investigated smartphone-based human’s decision making since the first generation of smart phones were available. Our results show that those who use smart phones tend to achieve “better” decisions to some extent. Second, we investigated how news sentiment may impact on human’s decisions. We examined how individuals responded to news sentiment by combining news sentiment from Thomson Reuters News Analytics over 10 years. Our finding news sentiment can alter individuals’ sequential decision-making. Third, we will discuss machine learning algorithms including deep neural networks that may cooperate with human’s decision-making for real-world decision-making applications.   I will informally discuss the challenges and lessons learned.