| Presentation | 1998/3/14 Memory-Based Reasoning using Correlational-Cross-Category Feature Importance and Quantification of its Classification Ability Tatsunori MATSUI, Akira YOKOZAWA, Toshio OKAMOTO, |
|---|---|
| PDF Download Page | PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | Memory-Based Reasoning (MBR) is a reasoning method for classification problem, in MBR knowledge base is constructed without rules for reasoning, only by keeping many cases represented by original vector form of features. In MBR, the weighting methods for features and its metric definition for similarity degree are important to get high accuracy. For example, weighting method based on conditional probability is defined in earlier studies. In this study, the weighting method regarding for correlation among some features, called Correlationnal-Cross-Category Feature Importance (CCCF), is defined as a new method. And its statistical characteristics are compared with other methods by experiments using artificial data sets. Moreover, an index to quantify the classification ability of weighting methods is defined. Experimental results and quantified index show that the classification ability of CCCF is higher than of other weighting methods. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Memory-Based Reasoning / Knowledge Base / Correlation / Similarity Function / Classification Ability |
| Paper # | |
| Date of Issue |
| Conference Information | |
| Committee | ET |
|---|---|
| Conference Date | 1998/3/14(1days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | |
| Chair | |
| Vice Chair | |
| Secretary | |
| Assistant | |
| Paper Information | |
| Registration To | Educational Technology (ET) |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Memory-Based Reasoning using Correlational-Cross-Category Feature Importance and Quantification of its Classification Ability |
| Sub Title (in English) | |
| Keyword(1) | Memory-Based Reasoning |
| Keyword(2) | Knowledge Base |
| Keyword(3) | Correlation |
| Keyword(4) | Similarity Function |
| Keyword(5) | Classification Ability |
| 1st Author's Name | Tatsunori MATSUI |
| 1st Author's Affiliation | Graduate School of Information Systems, The University of Electro-Communications() |
| 2nd Author's Name | Akira YOKOZAWA |
| 2nd Author's Affiliation | Graduate School of Information Systems, The University of Electro-Communications |
| 3rd Author's Name | Toshio OKAMOTO |
| 3rd Author's Affiliation | Graduate School of Information Systems, The University of Electro-Communications |
| Date | 1998/3/14 |
| Paper # | |
| Volume (vol) | vol.97 |
| Number (no) | 606 |
| Page | pp.pp.- |
| #Pages | 8 |
| Date of Issue | |