In data mining, which aims at sophisticated discovery of potentially useful and understandable patterns from massive data, we tackle diverse issues from fundamental ones to applications with various backgrounds including machine learning. Examples include data processing such as data squashing and data structure, pattern discovery such as various types of exceptions and rules, pattern interpretation such as information visualization and human factors, and other issues such as problem formalization. Moreover we conduct research on autonomous mobile robots using machine learning and data mining techniques.
Division of Bioinformatics
Bioinformatics is an interdisciplinary research field of biological and information sciences that was introduced during the genome project. In order to master the genome and its applications to medical fields, we need not only the basic biological approaches, but also information-oriented models and applications from basic to advanced levels. This research field focuses on education and research that enable students to analyze subjects from the genome to the basic principles of life on the basis of information theory and various types of informatics. For this purpose, our program provides graduates with cutting-edge knowledge about measurement theory, mathematical science, statistics, basic informatics, database, algorithms, machine learning, bioinformatics, neuroscience, and their applications to biomedical sciences.
■Data Mining and Bioinformatics
Professor Einoshin Suzuki
Associate Professor Hiroshi Yoshida
Tissues of animals and plants are maintained through balanced cell growth,movement, and elimination. Although cells are exchanged perpetually, the whole structure of the tissue is maintained. This form of maintenance is called cell turnover. I proposed a bio-inspired model of patterns that regenerate through turnover. This model is derived from the Dachsous–Fat system, which has recently attracted some attention because it is considered to facilitate regeneration in insect legs. In this model, I parameterize the manner of the redistribution of Dachsous and Fat during cell division, and then derived equations in the parameters that enable the patterns to regenerate and maintain themselves through turnover.Extending these models, I now propose a method for analyzing multicell-turnover models by using multivariate polynomials (Polynomial-life model).
■Neuroimaging and Neuroinformatics
Professor Keiji Iramina
Our laboratory specializes in non-invasive functional brain imaging, which includes the measurement of electroencephalograph (EEG), near-infrared spectroscopy (NIRS) and transcranial stimulation. Transcranial magnetic stimulation (TMS) and transcranial direct current stimulation (tDCS) is able to generate a long-term increase or decrease in the neuronal excitability that can modulate cognitive tasks non-invasively. We develop, integrate, and apply new ideas through innovative interdisciplinary research. The elucidation of the mechanisms of brain function is one of the foundations for the life sciences; it can be applied to almost all of its fields. In our studies we aim to obtain a deep understanding of brain information processing, and to apply the research results to life sciences and medicine. We also develop a brain computer interface and brain machine interface for welfare and education of disabled people.
Associate Professor Tsuyoshi Okamoto
We research the sensory systems in the brain (especially, vision, olfaction, and warmth). The aim is not only to clarify the mystery of the sensory systems. From our research findings, we would like to create something completely new (comfortable environments, effective teaching methods, research themes, and art works). For these objectives, we use three different scientific approaches: experiments (EEG, ERP, fMRI, psychophysical measures, etc.), data analyses (waveform analysis, image analysis, statistical analysis, etc.), and theories (mathematical model, computational model, numerical simulation, etc.).
Professor Seiichi Uchida
Faculty of Information Science and Electrical Engineering
Our lab tackles with automatic image analysis of bioimages, such as microscopic intracellular images. This is a very challenging research topic because of background noise, low resolution, low contrast, low frame-rate, blur, low appearance information, multiple target components, heavy nonlinear deformation, large variability, disappearance, etc. Our lab is developing new image processing and pattern recognition technologies to deal with such difficulties.
Associate Professor Kei Hirose
Institute of Mathematics for Industry (IMI)
Sparse Multivariate Analysis via L1 Regularization
Recently, the analysis of big data has becoming more and more important. Although the data volumes are increasing, most of the data values can be meaningless. Therefore, it is important to extract meaningful information from the big data. The sparse estimation, such as L1 regularization, is one of the most efficient methods to achieve this. The sparse estimation makes most of the parameters exactly zeroes. The meaningful variables correspond to the nonzero parameters. A remarkable feature of the L1 regularization is that even if the number of parameters is several millions, it takes only several minutes to compute the solution. In particular, I developed an efficient algorithm for computing the entire solutions for factor analysis, and made an R package fanc (https://cran. r-project.org/web/packages/fanc/index.html).
Professor Kenshi Hayashi
Faculty of Information Science and Electrical Engineering （Ito campus）
Main research subject is a development of functional devices based on wide variety of organic electronic materials; Au nanoparticles for low dimensional conductor, plasmonic nanoparticles, nanowire conductor, molecular wire, low dimensional materials and network system, optical sensing, molecularimprinting and aptamer nano-scale material and structure for molecular recognition. Chemical sensor devices such as odor sensor, odor-imaging devices for gas distributed chemical world, nano-scale organic electronic devices for high functional electronic systems are our research targets. Information processing of sensor output, biological odor cluster map image processing in biological olfactory bulb and odor database based on odor clustering are also our aims. Applications of the developed sensing system are biometrics and human detection, medical diagnosis, environmental sensing, agricultural information technology, odor source localization for various hazardous odor and fire detection.
■Gene Expression Control
Associate Professor Kosuke Tashiro
Molecular Gene Technology:
Regulatory networks function in the maintenance, adaptation and development of life. We focus on the transcriptional regulation mechanisms, and study the regulatory network in single-cellular and in multi-cellular organization. Our research topics include: the cellular gene regulation system using experimental and computational strategy, gene regulation in cellular differentiation, gene regulatory networks in strawberry, the development of devices for analysis of environmental microorganisms, and structural analysis of microorganism genome structure.
Professor Johan Lauwereyns
“I think therefore I am” (cogito ergo sum). Everyone knows these words by Descartes. His method was based on reasoning and doubting. Doubt came first. We can rephrase it as: “I doubt therefore I think therefore I am” (dubito ergo cogito ergo sum). We borrow this as a motto for our lab. Doubt comes first, as a scientific method and as subject for investigation. We wonder and inquire about perception and decision-making in complex or ambiguous situations. We study doubt with doubt.
Approach and objectives:
1.We develop a dynamical systems approach to analyzing neural data: Moving beyond the traditional static and rigidly deterministic perspectives in behavioral analysis and neurophysiology.
2.We look at improving the bioethical reasoning and praxis in neuroscience: Toward a scientifically and ethically motivated approach to reevaluating, revising and optimizing the use of animal models.
3.We investigate the cognitive dimensions of behavior (and their analogues) in various animal models, from rats to worms and back to humans.
4.We integrate cross-disciplinary perspectives in Psychology, Neuroscience, Philosophy, and Arts and Letters. This implies replacing psychoanalysis with cognitive neuroscience as the privileged partner to investigate matters of consciousness and the mind