Enhancing Sequential Access Efficiency in Data Compression: An Alpha-Parameterized Extension to MFLP for Dynamic Reorganization
Baisakh, Nileshkumar Patel, Hitesh Mohapatra, Sarthak Kalia · 2024
Online optimization problems throw some potential challenges in the process of designing and analyzing algorithms as input sequences are not known in advanced. Under such circumstances, how about the efficacy of deterministic algorithms to solve these problems in the online framework. This article has considered a well known online problem known as the list update problem which has been studied extensively in theoretically as well as experimentally to understand the behavior of various deterministic online list update algorithms. Building on the principle of dynamic list rearrangement, this paper introduces a novel alternative to traditional deterministic algorithms. Specifically, a alpha parameterized based Weighted Move-To-Front-Logarithmic Position W-MFLP algorithm is proposed. The performance of the new variant of MFLP algorithm is evaluated through experimental comparison with well-known deterministic algorithms, such as MTF and MFLP. The results show that the W-MFLP algorithm consistently outperforms MTF and MFLP, reinforcing the notion that adaptive approaches should be favored over deterministic ones when handling request sequences under uncertainty.