Bridging the Human Machine Gap

Humans have the remarkable ability to recognize an object by observation and experience or apply existing knowledge to solve a problem in an entirely new scenario. Such an innate sense of intuition is difficult to put into words, let alone code into an algorithm so that machines can emulate our behavior. Though achieving true artificial intelligence appears to be miles away, Prof Qiang Yang has been making substantial progress in building machine intelligence.

An expert in data mining, artificial intelligence, machine learning, transfer learning and deep reinforcement learning, Prof Yang has been developing algorithms that give computers similar capabilities to humans in retaining and reusing previously learned knowledge from mining large sets of data. His research has yielded him a first in integrating a reinforcement learning algorithm leveraging user feedback with transfer and deep learning so that machines can make more intelligent decisions. He and his team have managed to greatly enhance machine reading capabilities, such that they are able to produce high readability summaries of long reports, helping to save time in this era of information overload. They have even designed a machine that can write a high-quality novel in the style of an author, taking just a few seconds to do so.

Prof Qiang Yang is Chair Professor and Former Department Head of Engineering and Chair Professor of Department of Computer Science and Engineering at HKUST. Prof Yang is the President of Hong Kong Society of Artificial Intelligence and Robotics (HKSAIR), and President of the International Joint Conference on Artificial Intelligence (IJCAI). He is also the Founding Editor-in-Chief of both ACM Transactions on Intelligent Systems and Technology (TIST) and IEEE Transactions on Big Data.