Prof. Irina Perfilieva, CSc., Prof. h.c.
University of Ostrava, CZ
Speech title: Fuzzy Transforms for Dimensionality Reduction with Applications to Big Data
Abstract: The "Big Data" processing is the main challenge of contemporary science. Experimental sciences like biology or chemistry are facing an explosion of the data available from experiments. However, much of the data is highly redundant and can be represented using a smaller number of parameters without significant loss of information. The underlying mathematical technique making possible this type of representation is called dimensionality reduction.
A modern approach is based on Laplacian eigenmaps, that characterize the local neighborhood in terms of the measure of closeness. The optimization problem focused on a low-dimensional representation of the data set is formulated, and the cost function is introduced. The proposed solution is formulated in terms of eigenmaps of the graph Laplacian.
In fuzzy literature, the most relevant dimensionality reduction technique is fuzzy (F) -transform. It is based on a granulation of a domain (fuzzy partition) and gives a tractable image of an original data. The main characteristics concerning input data: size reduction, noise removal, invariance to geometrical transformations, knowledge transfer from conventional mathematics, fast computation.
In the talk, we discuss how the F-transform can be explained in terms of Laplacian eigenmaps and combined with the PCA.
We demonstrate the efficiency of the proposed combination on the example of pattern recognition in a large database. We also compare this combination with other relevant techniques (besides other, LENET-like CNN) from the computation time and success rate points of view.
Bio.: Professor Irina Perfilieva, Ph.D., received the degrees of M.S. (1975) and Ph.D (1980) in Applied Mathematics from the Lomonosov State University in Moscow, Russia. At present, she is full professor in the University of Ostrava (CZ) and a Head of Theoretical Research Department in the Institute for Research and Applications of Fuzzy Modeling. She is: the author and co-author of six books on mathematical principles of fuzzy sets, fuzzy logic and their applications; an editor of many special issues of scientific journals. She has published over 270 papers in the area of MV logic, fuzzy logic, fuzzy approximation, topology, image processing.
She is: area editor of IEEE TFS and International Journal of CIS, editorial board member of the journals: Fuzzy Sets and Systems, Journal of Uncertain Systems, Journal of Intelligent Technologies and Applied Statistics, Fuzzy Information and Engineering. She works as a member of Program Committees of the most prestigious International Conferences and Congresses in the area of fuzzy and knowledge-based systems. She received the memorial Da Ruan award at FLINS 2012 and the best paper award at IFSA 2019. She is an EUSFLAT Honorary Member, and the IFSA Fellow. She got a special price of the Seoul International Inventions Fair 2010. She has two patents.
Her scientific interests lie in the area of applied mathematics and mathematical modeling where she successfully uses modern as well as classical approaches. During last five years she is working in the area of image processing and pattern recognition.
Dr. Ghalib Alshammri
Computer Science, Community College – King Saud University, KSA.
Electrical & Computer Engineering – Stevens Institute of Technology, USA.
Speech Title: Artificial Intelligence Techniques for Diffusion-Based Bio-Molecular Nano Communication Networks
Abstract: Molecular communications (MC) is an emerging field that is sought to serve in many vital future nanotechnology applications such as drug delivery and nanomedicine. In this talk, we tackle two challenging problems in MC System and network design, namely, data detection at the link level and optimal relay design at the network level.
Threshold-based detection and cooperative (relay) transmission techniques have been developed for wireless sensor networks (WSNs) in recent years to improve network transmission rate, delay and bit error rate (BER) performance, while not as much research has taken place for molecular communication-based systems. The special characteristics and constraints of bio-molecular nano communication network, such as dynamic environmental conditions and various nanonetwork topologies, makes it necessary to revisit the techniques used for threshold-based detection and cooperative relay transmission. Therefore, we propose detection approaches that consider and rake into account the various MC channel parameters (e.g., radius of the propagating molecules, viscosity, drift velocity, and the temperature of the fluid environment). We also consider and propose novel solutions for the challenging problem of inter-symbol interference (ISI) and reception noise sources.
This talk provides two key contributions: First, we propose and investigate high performance receiver designs using novel threshold-based detection techniques for On-OFF-Keying diffusion-based bio-molecular communication nanonetwork systems. Techniques we contribute include:
1-Molecular Memory-Assisted Threshold-Based Detection Technique.
2-Fuzzy Rule-Based Classification System.
3-NFIS: Neuro-Fuzzy Inference System using Polynomial Approximation.
4-Low-Complexity Machine Learning Detection Technique.
5-ARFIS: Adaptive-Receiver-Based Fuzzy Inference System. These techniques achieve very low incomes bit error rates compared to traditional approaches.
Second, we propose and investigate a long-range nanoscale MC network system based on intelligent multi-agent decision support mechanisms that aggregate the sensed data and simplifies the complexity of the relay techniques, where the intermediate intelligent agents do not decode the received/related information. Our proposed techniques provide a novel cooperative decision system that utilizes intelligent agents in combination with pulse energy ratio and sense-and-release technique. We evaluate the performance of our proposed techniques various system parameters, such as diffusion coefficients and transmission distance.
Bio.: Ghalib Alshammri received the B.Ed. degree in computer science (distinction with honor) from King Saud University, Riyadh, Kingdom of Saudi Arabia, in 2003 and the master’s degree in computer science (Major: Intelligent Systems) from University of Wollongong, Australia, in 2009. Also, he received the Ph.D. degree in electrical and computer engineering from Stevens Institute of Technology, Hoboken, NJ, USA, on May 2019. During 2003 – Now he is a lecturer at the department of Computer Science, Community College, King Saud University. During Spring 2017 – Now he is a Research and Teaching Assistant at the Electrical and Computer Engineering – Stevens Institute of Technology. Dr. Ghalib has received numerous major scholarships and grants including Saudi Arabian Cultural Mission (SACM), National Science Foundation (NFS) and Graduate Student – Stevens Institute of Technology. Dr. Ghalib’s research interests and technical experience include information theoretic considerations, such as coding, signal processing, signal detection and estimation, wireless multiple-access techniques, transceiver design. molecular communication nanonetworks, Internet of Bio-Nano Things (IoBNT), Artificial Neural Network (ANN), Fuzzy concept, machine learning, robotics, and agent techniques. Dr. Alshammri holds U.S. provisional patent to-date and has authored more than 10 original research papers. He is a reviewer in numerous major conferences and journals; e.g., IEEE International Conference on Communications (ICC). During 2013 – 2014 he works as consultant in general department of certificates equalization, Ministry of Education (MoE), Kingdom of Saudi Arabia. Finally, Dr. Alshammri is teaching the numerous of CS and ECE courses and supervising graduation projects for undergraduate students at King Saud University and Stevens Institute of Technology.