Dr. Suresh Sundaram
This talk present recently developed a supervised learning algorithm for a spiking neural network. In general, a spiking neural network closely emulates the structural and behavioral properties of the biological brain. This spiking neural network property motivated researchers to develop an advanced learning mechanism closely mimic the human brain. The discontinuous nature of spike and dynamic plasticity of synapse makes it difficult to extend the existing learning principles to spiking neural networks. In this talk, we present two aspects of the spiking neural network. 1) evolving spiking précising spike learning rule, which exhibits both local and global learning behavior. The learning algorithm is adaptive to the synaptic weight based on modulation factor from meta neuron and desired change in postsynaptic potential required. Since the learning algorithm captures the information in the one-shot, an online meta neuron-based learning algorithm evolves the structure and simultaneously adapts the parameter. 2) Time-varying long-term Synaptic Efficacy Function-based leaky-integrate-and-fire neuRON model, referred to as SEFRON and its supervised learning rule for pattern classification problems. The time-varying synaptic efficacy function is represented by a sum of amplitude-modulated Gaussian distribution functions located at different times. Similar to the gamma-aminobutyric acid-switch phenomenon observed in a biological neuron that switches between excitatory and inhibitory postsynaptic potentials based on the physiological needs, the time-varying synapse model allows the synaptic efficacy (weight) to switch signs in a continuous manner. Finally, we highlight the interpretability of the spiking neural network.
Suresh Sundaram is an associate professor in Department of Aerospace Engineering, Indian Institute of Science, Bangalore, India from 2018. He is leading the WIPRO-IISc joint research initiative on Artificial Intelligence driven Autonomous System. Prior to the current appointment, he was a faculty in School of Computer Science and Engineering, Nanyang Technological University, Singapore between 2010-2018. During the period, he served as a cluster director in Energy Research Institute @ NTU and deputy director at Robotics Research Center, NTU.
Dr. Suresh received his bachelor of engineering from the Bharathiar University in 1995 and his master in engineering (2001) and Ph.D from Indian Institute of Science in 2005. He has been associate editor in IEEE Transactions on Neural Networks and Learning System, Swarm and Evolutionary Computing and Evolving System. He was the general chair for IEEE Symposium Series on Computational Intelligence in 2018 and Publication chair in 2013. He published a book and more than 300 articles. His research interest include intelligent flight control, autonomous system, applied game theory and artificial intelligence.
This talk aims to present an overview of the common pitfalls for applying machine learning techniques to real-world problems from a perspective of fairness. This talk mainly highlights the importance of diversity of the data and the problem related to algorithmic bias. In the age of information overload, machine learning becomes increasingly important in everyday life. There has been a growing interest in discovering the harmful effect of bias in machine learning and a way to take fairness into service. Based on our research and experience in the industry, we discuss open questions for further application.
Tomoki Fukuma is the founder and CEO of TDAI Lab, a machine learning AI startup, founded in Tokyo in 2016. He is now a Ph.D. student in the Department of Systems Innovation, School of Engineering, The University of Tokyo. The main area of his interest is learning the true quality of the content in an online platform. This includes topics in unbiased learning-to-rank, recommendation, and AI for Society. He is also a Japanese ballroom dancer, representing Japan from 2016 to 2019.