Intelligent decision: towards interpreting the Pe Algorithm

TL;DR

This paper introduces the Pe algorithm, combining RDD and IIR algorithms, to interpret intelligent decision-making through outlier detection, providing clear definitions, efficient methods, and experimental validation.

Contribution

It presents a novel Pe algorithm that unifies outlier detection and decision-making, with clear definitions and an example application to curve outliers.

Findings

Pe algorithm demonstrates robustness in experiments

Unified approach improves outlier detection accuracy

Provides theoretical and practical insights into intelligent decision-making

Abstract

The human intelligence lies in the algorithm, the nature of algorithm lies in the classification, and the classification is equal to outlier detection. A lot of algorithms have been proposed to detect outliers, meanwhile a lot of definitions. Unsatisfying point is that definitions seem vague, which makes the solution an ad hoc one. We analyzed the nature of outliers, and give two clear definitions. We then develop an efficient RDD algorithm, which converts outlier problem to pattern and degree problem. Furthermore, a collapse mechanism was introduced by IIR algorithm, which can be united seamlessly with the RDD algorithm and serve for the final decision. Both algorithms are originated from the study on general AI. The combined edition is named as Pe algorithm, which is the basis of the intelligent decision. Here we introduce longest k-turn subsequence problem and corresponding solution as an example to interpret the function of Pe algorithm in detecting curve-type outliers. We also give a comparison between IIR algorithm and Pe algorithm, where we can get a better understanding at both algorithms. A short discussion about intelligence is added to demonstrate the function of the Pe algorithm. Related experimental results indicate its robustness.