LED-Based Adaptive Algorithm: A Novel Integration of Logit and Exponential Decay for Self-Organizing Sequential Search Problem

This research presents an innovative adaptive list reorganization strategy aimed at minimizing the total access cost for dynamic sequences. The proposed algorithm leverages a dual-parameter approach combining Logit and Exponential Decay (LED) model for the self-organizing linear list problem, also known as the list update problem. The proposed method presented in this article, is one of the few existing approaches that offers an alternative to deterministic strategies, specifically designed to handle random input sequences for this problem.This strategy significantly improves performance compared to deterministic algorithms like Move-to-Front (MTF) and Move-to-Middle (MTM), which fail to account for varying patterns of locality and frequency in request sequences. The comparative analysis is performed with the well-known deterministic list update algorithms such as MTF and MFM by considering the Calgary Corpus standard dataset. The experimental results show that the adaptive reorganization strategy outperformed both algorithms.