Increasing Accuracy and Interpretability of High-Dimensional Rules for Learning Classifier System
| Autor: | Masakazu Tadokoro, Hiroyuki Sato, Hiroki Shiraishi, Yohei Hayamizu, Keiki Takadama, Yukiko Fukumoto |
|---|---|
| Rok vydání: | 2021 |
| Předmět: |
Learning classifier system
business.industry Computer science Sampling (statistics) Pattern recognition Evolutionary computation ComputingMethodologies_PATTERNRECOGNITION Encoding (memory) Classifier (linguistics) Reinforcement learning Artificial intelligence business Decoding methods Interpretability |
| Zdroj: | CEC |
| Popis: | This paper proposes ELPSDeCS (Encoding, Learning, "Plausible" Sampling, and Decoding Classifier System) by extending ELSDeCS (Encoding, Learning, Sampling, and Decoding Classifier System) to increase both the accuracy and interpretability of the generated classifiers which matches the high-dimensional input such as images. The experimental results on the complex multi-class classification problem of the handwritten numerals show that both the accuracy and interpretability of ELPSDeCS are higher than that of ELSDeCS. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|