ide-research.net
Top | Publications | Project  [Japanese/English]

Tsuyoshi Ide's Home Page

This page introduces the projects I joind (or, am working on). Since around 2003, I have engaged in several data mining projects at IBM Research. Currently, I'm working as a sort of data analytics consultant, focusing on knowledge discovery techniques from real-valued data generated in manufacturing industries. I believe that our activities are at the frontier of service science.


Current project

Since the end of 2010, I have been leading Analytics & Optimization at IBM Research - Tokyo. I have defined two strategic research areas:

For the analysis of stochastic interacting systems, our ultimate goal is to establish the methodology to analyze complex systems such as societies, cities and enterprises. Examples of our research include analysis of transportation traffic and human decision making in the market.

Since establishing fully analytic models is hopeless in complex systems, simulation technologies can be a powerful approach. However, one critical issue is how to validate the simulation result. To address this, we are interested in how simulation is combined with optimization technologies. For example, we may want to optimize the model of individual agents in multi-agent traffic simulation using sophisticated machine learning technologies. This type of problem is actively explored topic in my team, and I am leading a Strategic Initiative in IBM Research in this area.

For the analysis of industrial dynamic systems, major research topics include sensor data analytics and production optimization, which are frequently addressed problems especially in the Japanese market. It is my pleasure to observe that much of my research including anomaly detection is playing a critical role in this category.


Previous projects

Sensor Data Analytics (2006-)

After the Autonomic Computing project, I launched a new project, Data Analytics for Quality Control, or simply Sensor Data Analytics, which aimed at improving the quality of products mainly in the manufacturing industries by taking full advantages of advanced analytics for sensor data. Since then I have involved in many customer projects as on-demand innovation services.

When I initiated the project, utilizing the data simply meant keeping track of parts traceability. However, now people seem to be getting to understand how advanced analytics works to change the standard way of business. This is the idea behind the Smarter Planet campaign and Business Analytics and Optimization (BAO). 

From a research perspective, the establishment of the notion of correlational anomaly detection is perhaps the most important contribution of mine. In the SDM paper above, I first introduced the sparse graphical model in the context of correlational anomaly detection. 


Automated Analysis Initiative (AAI; 2003-2005)

At least in the Tokyo Research Lab., the autonomic computing project was not very successful. In my opinion, the activities looked more like just development work, and lacked original research agenda. This project, a joint effort between Watson and Tokyo, aimed at developing a general framework for sensor data particularly in the automotive industry. I introduced the new notion of change-point correlation (the SDM paper), which became an important part of the framework. My work was to address the heterogeneity over different sensor data, which is a common nature in industrial sensor data. The success of this attempt motivated my next project, Sensor Data Analytics.


Autonomic Computing (2002-2004)

This project was a company-wide initiative that aimed at handling the growing complexities of computer systems. While I had just started in this new area, having moved from the totally different area of LCD technologies, I set a research agenda that would be useful in the domain: anomaly detection for system monitoring. The KDD paper, which is my very first paper in computer science, was written in this project.


Collimated backlight (2000-2001)

The goal of this project was to develop a new type backlight system that had the world-highest light utilization efficiency. The new design, “collimated backlight,” had been introduced for that, but it suffered from a serious problem in practice: visible more patterns on the display, which is caused by the optical interference between a light guide and a grid-like circuit pattern of LCDs. I invented a novel technology to clearly solve this critical issue. The idea was to develop specially-designed random dot patterns. I proposed a totally new solution that uses low-discrepancy sequences combined with a molecular dynamics model.