IN 2016, there was a news report that made headlines in the media. Guess what? None other than Lee Sedol, who was one of the champions of the strategy board game called Go, was defeated by Google’s DeepMind AlphaGo!
Over the next few years, a host of champions had also lost in the game.
Since then, artificial intelligence (AI), particularly the variants of deep learning such as deep reinforcement learning (DRL) used in AlphaGo, has received great attention with researchers and developers racing to create smart systems using artificial intelligence (AI), or to turn smart systems into even smarter systems, in the hope of achieving better-than-human intelligence in a diverse range of systems, such as driverless cars and industry automation. Ultimately, deep reinforcement learning (DRL) contributes to achieving sustainable
To mimic the human brain structure, deep learning uses several layers of artificial neurons to store the learned knowledge.
For deep reinforcement learning (DRL), the agent (or decision maker) explores the potential actions, and exploits the best possible action under different states of operating environment.
Exploration is essential because of the dynamic operating environment (or state). Meanwhile, exploitation helps to achieve the highest possible action during exploitation.
Ultimately, the agent chooses exploitation actions most of the time to maximise the accumulated rewards.
Compared to reinforcement learning, deep reinforcement learning (DRL) uses a deep neural network to represent a complex set of states and has been shown to achieve breakthrough performance with lower computational cost, reduced learning time, and more efficient knowledge storage.
Compared to the traditional reinforcement learning approach, deep reinforcement learning (DRL) represents complex states and has been shown to achieve breakthrough performance with lower computational cost, reduced learning time, and more efficient knowledge storage.
UTAR’s Sungai Long campus offers the Bachelor of Science (Honours) Software Engineering as a three-year programme after foundation/STPM/UEC/diploma.
This programme equips students with both fundamental and advanced features of traditional artificial intelligence (AI) approaches and deep learning.
Our students are also equipped with knowledge of how to use the learned approaches to solve real world problems.
Our programme is designed with input from the software and IT industry and benchmarked to the best universities around the world. Bachelor of Science (Honours) Software Engineering is accredited by Malaysian Board of Technologist (MBOT).
To learn more about programmes offered by LKCFES, East Malaysia Education Fairs and Virtual Open Day, please visit study.utar.edu.my or call 016-2233 557.