|Systems Ai Lab.||Intelligent informatics, sensor systems, game theoretical approach to agents, diagrammization|
|Intelligent informatics, sensor systems, game theoretical approach to agents, diagrammization|
The Systems Science Laboratory studies operational principles of systems from the intelligence of life/nature/society. On the basis of these principles, we aim for design principles of resilient systems by system modeling, simulations, analysis, and synthesis. Artificial intelligence with system theoretical approaches, as well as agent technologies including swarm intelligence, are the topics focused on. We also study complex systems, immune informatics, and bio intelligence.
On the basis of self-recognition/repair/reconfiguration models and simulations, we design and build servers resilient against physical damage as well as malicious acts.
Formation of many agents by autonomous distributed algorithms
For the swarm intelligence of a group of agents, we develop autonomous distributed algorithms for the agent.
Disaster reduction simulators and resilient systems for ICT
On the basis of self-recognition computers and autonomous distributed systems, we design disaster reduction simulators and resilient systems for ICT.
Solar power systems simulation by cellular automaton
We study and design simulations for solar power systems based on a cellular automaton technique.
|Discrete Optimization Lab.||Algorithms, combinatorial optimization, mathematical programming, online models, constraint satisfaction problem|
|Algorithms, combinatorial optimization, mathematical programming, online models, constraint satisfaction problem|
Computational tasks required by modern society and industry vary widely. It is however often observed that naively designed algorithms give rise to exponential explosions in their running costs and poor-quality outcomes, and software development for highly complicated tasks are known to be almost matters of algorithm development. The goal of our research is set up accordingly to design and develop algorithms of high efficiency and accuracy for various types of computational problems, with the help of discrete structure analysis, algorithm theory, complexity theory, mathematical programming, and others.
Design and development of new algorithms and modelings
Among hte lab's goals is to design new algorithms and/or mathematical models to deal with various types of important combinatorial optimization problems, typically modeled as network/ graph problems, arising mainly in production/transportation planning, scheduling, VLSI design, and optimal routing. The validity and efficiency are evaluated by theoretical analysis and/or computer experiments.
Development of algorithm design techniques
This initiative is intended to study optimization principles and mechanisms, by a level of consideration higher than the previous theme, inherent in algorithms for combinatorial optimization as driving forces, such as design techniques based on the duality theorem and the complementary slackness in linear programming.
It is required, in online problems, to finish processing one datum before starting the next, where input data are provided over time; such problems can be found in a wide range of applications from power-saving control to logistics and financial engineering. Moreover, it is of greater importance, in the era of big data, to design streaming algorithms for processing large-scale sequential data under severely controlled memory storage.
|Computers and Education Lab.||Information education, language teaching, computers andeducation, e-learning, HCI|
|Information education, language teaching, computers andeducation, e-learning, HCI|
National curriculum standards reform for primary and secondary schools were announced in 1998. Consistent and systematic information education through all stages of school education require sufficient improvement of related subjects and the active use of computers. The author collaborated with teachers of adjacent primary and secondary schools on establishing new teaching materials for information education.
Information Education using Computers
Language Teaching using Computers
Practical Study on Science-Cafe
|Quantum biology Lab.||Molecular simulation, quantum science, electronic states calculation,DNA, protein, transcription mechanism of DNA genome information,inhibitor to cancer metastasis, medicine for Alzheimer's disease,tuberculosis drug, artificial nucleobase|
|Molecular simulation, quantum science, electronic states calculation,DNA, protein, transcription mechanism of DNA genome information,inhibitor to cancer metastasis, medicine for Alzheimer's disease,tuberculosis drug, artificial nucleobase|
We have been studying stable structures and electronic states of DNA, protein, and ligands and their complexes, using our developed molecular simulations and parallel computers. Our final goal is to elucidate the transcription mechanism of genome information from DNA and RNA controlled by many kinds of transcriptional proteins. In addition, we would like to propose novel compounds that efficiently control the functions of proteins related with diseases. The main topics of our recent study are as follows.
Molecular simulations for elucidating the transcription mechanisms of genome information controlled by transcriptional proteins and ligands
Proposal for novel medicines for cancer metastasis based on ab initio molecular simulations
Proposal for novel medicines for Alzheimer's disease that inhibit the generation and the aggregation of amyloid-beta peptides
Proposal for novel agents to inhibit the function of proteins controlling the pathogenesis of tuberculosis and the growth of tuberculosis bacteria
Molecular simulations of the specific interactions between dioxin and xenobiotic receptor protein for elucidating the mechanism of the onset of toxicity induced by dioxin
Molecular simulations of the charge transfer mechanism through the duplexes composed of artificial nucleobases for proposing a novel biochip alternative to the DNA chip
|Computational Chemistry Lab.||Computational material and molecular science, High-performance computing|
|Computational material and molecular science, High-performance computing|
Although the elucidation of phenomena caused by invisible small molecules is one of the most difficult problems, those consequently becomes the very challenging research themes. In order to investigate their mysterious and interesting material and molecular science, our computational chemistry laboratory is studying on the developments of computational simulation and visualizing applications based on the experiments and theory of natural sciences (physical, chemical and biological and mathematical knowledge) and the latest informatics technologies.
To realize the drug-discovery and biological phenomena analysis, and to represent the high-accurate prediction of nano/bio materiel properties, we are working on the developments of various high-performance computation tools; for example, a molecular simulation application "CONFLEX" based on the computational chemistry and bioinformatics, 3D molecular/material visualization tool "BARISTA", etc.
Recently, we challenge the various studies on new medicines of flu, cancer and Alzheimer's diseases, and also new functional materials of power fuel cell, semiconductor, 3D-display resources.
|Computer Systems Performance Engineering Lab||Computer Architecture, Computer Systems, Software Performance Engineering|
|Computer Architecture, Computer Systems, Software Performance Engineering|
The Computer Systems Performance Engineering Laboratory studies performance of any platforms of computers from ultra high-speed supercomputers to extreme low-power mobile devices based on scientific and engineering approaches. Currently, many deep learning programs, which become a typical component for AI-based applications, has struggled with lack of performance even if it is implemented on the state-of-the-art CPUs or GPUs. Therefore, techniques that realize high-performance and high-efficiency computer systems are expected to be an enabler of emerging new AI applications such as self-driving cars and autonomous intelligent robots. In the context of system performance matter, we highlight specialization for inherent memory access locality which will be a clue to its solution.
Memory-centric customization and co-design methodology
Scientific modeling of system performance and quality of the performance
Automation of customization and co-design driven by mathematical optimization and machine learning techniques