Bor-Shyh Lin

Professor

Institute of Imaging and Biomedical Photonics

National Chiao Tung University, Taiwan

Title: Biosensing Technology and Brain Computer Interface Design

Abstract:

Brain–computer interface (BCI), is a direct communication pathway between the brain and an external device. BCIs are often directed at assisting, augmenting, or repairing human cognitive or sensory-motor functions. However, in reality, there are a lot of technical problems coming from brain physiology, in particular, noninvasive sensing techniques for brain activity and partial brain knowledge. Most of current BCI systems require the conventional EEG machine and EEG electrodes with conductive gels (wet electrodes) to acquire multi-channel EEG signals, and then transmit these EEG signals to the back-end computer to perform the function of brain computer interface. This also reduces the convenience of use in daily life, and increases the limitation of BCI applications. In this talk, we introduce the basic scheme of brain computer interface and several novel sensing techniques for brain activity, including novel dry electrodes and near-infrared spectroscopy. Different BCI applications are also illustrated and demonstrated.

Biosketch:

Bor-Shyh Lin, Ph.D., is Professor of Institute of Imaging and Biomedical Photonics, National Chiao Tung University, Taiwan. He is also IEEE senior member. He received B.S. degree in electrical engineering from National Chiao Tung University, Taiwan, in 1997, M.S. degree and Ph. D. degree in electrical engineering from National Taiwan University, Taiwan, in 1999 and 2006 respectively. He established and founded Biomedical Systems Laboratory in 2009 and Chi Mei Medical Center-National Chiao Tung University Biomedical Engineering Joint Laboratory in 2013. His research interests are in the areas of biomedical system design, biomedical signal processing, and biosensor. He is also the recipient of several academic awards including IEEE Best GOLD Member Award.

Sebastián Basterrech

Department of Computer Science

Faculty of Electrical Engineering

Czech Technical University, Prague, Czech Republic

Title: An overview on Randomized Neural Networks

Abstract:

During the last years, there have been an increasing interest in randomized Neural Networks (NNs). This NN family is pretty large, some of the most popular methods include: drop-out techniques, random projections, Random Neural Networks, Extreme Learning Machines and Reservoir Computing methods. The models present several advantages among them we can mention their good performance for solving well-known benchmarks, their generalization ability, easy implementation, the modularity in their design, as well as their biological plausibility.
In this talk, we present an overview of randomized NN techniques, we cover RC, ELM, and deep NNs. Besides, we present a new technique named Echo State Queueing Network (ESQN) that combines characteristics of RC techniques and good properties of queueing networks. We discuss about the problems of learning instability, the advantages and limitations of randomized NNs. Furthermore, we discuss applications of soft-computing algorithms for improving the performance of randomized NNs. We present several empirical results on simulated and real time-series data.

Biosketch:

Sebastián Basterrech obtained his Ph.D. degree in Computer Sciences in November, 2012 from the University of Rennes 1, Bretagne, France. During 2009-2012 he was a doctoral student fellowship of INRIA, France. During two years he has been working as a postdoctoral researcher at the National Supercomputing Center of Czech Republic. During 2016, he was part of the Computer Department at Faculty of Electrical and Computer Science in Technical University of Ostrava. Currently, he is researcher at the Department of Computer Science in Faculty of Electrical Engineering, Technical University of Prague. S. Basterrech is a TC member of the International IEEE SMC Soft-Computing Society, and he is member of the Neural Network Society. He has several contributions in areas related to quasi-Newton optimization, Random Neural Networks, Reservoir Computing, Neural Computation and Soft-computing techniques.