Muneaki Ohshima

Peculiarity oriented fMRI brain data analysis for studying human multi-perception mechanism

Although many cognitive and brain scientists have already studied human information processing mechanism of auditory and visual, separately, the relevance between auditory and visual information processing needs to be investigated in depth. We investigate human multi-perception mechanism by combining various psychological experiments, physiological measurements, data cleaning, modeling, transformation and mining techniques for developing artificial systems which match human ability in specific aspects. In particular, we observe fMRI (functional magnetic resonance imaging) brain activations from the view point of peculiarity oriented mining and propose a way of peculiarity oriented mining for knowledge discovery in multiple human brain data. We present an interesting result on modeling, transforming, and mining multiple human brain data obtained from visual and auditory psychological experiments by using the fMRI. The proposed methodology attempts to change the perspective of cognitive scientists from a single type of experimental data analysis towards a holistic view.

Peculiarity Oriented Mining and Its Application for Knowledge Discovery in Amino-Acid Data

The paper proposes a way of peculiarity oriented mining and its application for knowledge discovery in the amino-acid data set. We introduce the peculiarity rules as a new type of association rules, which can be discovered from a relatively small number of peculiar data by searching the relevance among the peculiar data. We argue that the peculiarity rules represent a typically unexpected, interesting regularity hidden in the amino-acid data set.

Interestingness, peculiarity, and multi-database mining

In order to discover new, surprising, interesting patterns hidden in data, peculiarity oriented mining and multidatabase mining are required. In the paper, we introduce peculiarity rules as a new class of rules, which can be discovered from a relatively low number of peculiar data by searching the relevance among the peculiar data. We give a formal interpretation and comparison of three classes of rules: association rules, exception rules, and peculiarity rules, as well as describe how to mine more interesting peculiarity rules in multiple databases.

Relational peculiarity-oriented mining

Peculiarity rules are a new type of useful knowledge that can be discovered by searching the relevance among peculiar data. A main task in mining such knowledge is peculiarity identification. Previous methods for finding peculiar data focus on attribute values. By extending to record-level peculiarity, this paper investigates relational peculiarity-oriented mining. Peculiarity rules are mined, and more importantly explained, in a relational mining framework. Several experiments are carried out and the results show that relational peculiarity-oriented mining is effective.

Peculiarity Oriented Analysis in Multi-people Tracking Images

In the place in which many people gather, we may find a suspicious person who is different from others from a security viewpoint. In other words, the person who takes a peculiar action is suspicious. In this paper, we describe an application of our peculiarity oriented mining approach for analysing in image sequences of tracking multiple walking people. A measure of peculiarity, which is called peculiarity factor, is investigated theoretically. The usefulness of our approach is verified by experimental results.

Relational peculiarity oriented data mining

Peculiarity rules are a new type of interesting rules which can be discovered by searching the relevance among peculiar data. A main task of mining peculiarity rules is the identification of peculiarity. Traditional methods of finding peculiar data are attribute-based approaches. This paper extends peculiarity oriented mining to relational peculiarity oriented mining. Peculiar data are identified on record level, and peculiar rules are mined and explained in a relational mining framework. The results from preliminary experiments show that relational peculiarity oriented mining is very effective.

Spiral discovery of a separate prediction model from chronic hepatitis data

In this paper, we summarize our endeavor for spiral discovery of a separate prediction model from chronic hepatitis data. We have initially proposed various learning/discovery methods including time-series decision tree, PrototypeLines, and peculiarity-oriented mining method for mining the data. This experience has motivated us to model physicians as considering typical cases with the specific disease and ruling out clearly exceptional cases. We have developed a spiral discovery system which learns a prediction model for each type of cases, and obtained promising results from experiments.

MULTI-ASPECT MINING IN HEPATITIS DATA

When therapy using interferon medication for chronic hepatitis patients, various conceptual knowledge/rules will benefit for giving a treatment. The paper describes our work on cooperatively using various data mining agents including GDT-RS, learning with ordered information (LOI), and peculiarity oriented mining (POM) in a spiral discovery process with the multi-phase such as pre-processing, rule mining, and post-processing, for multi-aspect analysis of the hepatitis data and meta learning. GDT-RS is an inductive learning system for discovering decision rules. LOI discovers ordering rules and important features. POM finds peculiarity data/rules. Our methodology and experimental results show that the perspective of medical doctors will be changed from a single type of experimental data analysis towards a holistic view, by using our multi-aspect mining approach.

Peculiarity Oriented Mining in Multiple Human Brain Data

In the paper, we investigate fMRI brain activations from the view point of peculiarity oriented mining and propose a way of peculiarity oriented mining for knowledge discovery in multiple human brain data. The mining process is a multi-step one, in which various psychological experiments, physiological measurements, and data mining techniques are cooperatively used to investigate human multi-perception mechanism. We describe the initial results on transforming the multiple human brain data obtained from visual and auditory psychological experiments by the functional magnetic resonance imaging (fMRI) as well as using peculiarity oriented mining technique in such multi-data.

Multi-strategy instance selection in mining chronic hepatitis data

In this paper, we propose a method which splits examples into typical and exceptional by mainly assuming that an example represents a case. The split is based on our previously developed data mining methods and a novel likelihood-based criterion. Such a split represents a highly intellectual activity thus the method is assumed to support the users, who are typically medical experts. Experiments with the chronic hepatitis data showed that our proposed method is effective and promising from various viewpoints.