Comments on "Fractional Extreme Value Adaptive Training Method: Fractional Steepest Descent Approach"

TL;DR

This paper critically examines a fractional steepest descent algorithm, highlighting overestimated convergence rates and a fundamental flaw that limits its effectiveness in adaptive learning applications.

Contribution

It provides a detailed critique of the original algorithm's convergence analysis and identifies a critical flaw affecting its practical applicability.

Findings

The convergence rate estimate is overstated.

A fundamental flaw hampers the algorithm's broad applicability.

Experimental results support the critique.

Abstract

In this comment, we raise serious concerns over the derivation of the rate of convergence of fractional steepest descent algorithm in Fractional Adaptive Learning (FAL) approach presented in `Fractional Extreme Value Adaptive Training Method: Fractional Steepest Descent Approach' [IEEE Trans. Neural Netw. Learn. Syst., vol. 26, no. 4, pp. 653--662, April 2015]. We substantiate that the estimate of the rate of convergence is grandiloquent. We also draw attention towards a critical flaw in the design of the algorithm stymieing its applicability for broad adaptive learning problems. Our claims are based on analytical reasoning supported by experimental results.