Online Learning Algorithm for Multi-Kernel Balancing of Computational Work Load

Authors

  • Ankush Rai Department of Applied Science, CRIAD Laboratories, Bhilai, Chhattisgarh, India

DOI:

https://doi.org/10.37591/jowet.v1i3.277

Abstract

The online learning is an advancement made toward the extended machine learning approach where the application is web-based. Thus, the common problem to build a kernel based online perceptron algorithm is the amount of memory required to dynamically access for the so formed hypothesis derivable from the algorithm. As it tends to increase without a lower bound; also the computational load of such an algorithm increases linearly with the memory chunk used to build the hypothesis. Though, the previous works are focused towards discarding the instances of online hypothesis from the perceptron algorithm in order to reserve the computational work load within the bounded memory. Therefore, in this study we presented a work which dynamically creates an Online Chain Reaction Algorithm (OCRA) for keeping the checkpoint trees of hypothesis with the varying memory instances; which makes the online learning more dynamics and invariant of the pre-requisite amount of memory required.

Keywords: online learning, kernel methods, distributed computation workload

References

Freund Y., Schapire R. E. Large Margin Classification using the Perceptron Algorithm. Mach Learn 1999; 277–296p.

Crammer K., Singer Y. Ultraconservative Online Algorithms for Multiclass Problems. J Mach Learn Res 2003; 3: 951–991p.

Crammer K., Kandola J., Singer Y. Online Classification on a Budget. In Adv Neural Info Process Syst 2003; 16: 225–232p.

Crammer K., Dekel O., Keshet J., et al. Online Passive-aggressive Algorithms. J Mach Learn Res 2006; 7: 551–585p.

Kivinen J., Smola A., Williamson R. Online Learning with Kernels. IEEE T Signal Process 2004; 52(8): 2165–2176p.

Weston J., Bordes A., Bottou L. Online (and offline) on an Even Tighter Budget. In Robert G. Cowell and Zoubin Ghahramani, editors, Proc. of AISTATS 2005, 2005, 413–420p.

Cheng L., Vishwanathan S. V. N., Schuurmans D., et al. Implicit Online Learning with Kernels. In Adv Neural Info Process Syst 2007; 19: 249–256p.

Dekel O., Shalev-Shwartz S., Singer Y. The Forgetron: A Kernel-based Perceptron on a Budget. SIAM J Comput 2007; 37(5): 1342–1372p.

Cesa-Bianchi N., Conconi A., Gentile C. On the Generalization Ability of On-line Learning Algorithms. IEEE T Inform Theory 2004; 50(9): 2050–2057p.

Cesa-Bianchi N., Conconi A., Gentile C. A Second-order Perceptron Algorithm. SIAM J Comput 2005; 34(3): 640–668p.

Cesa-Bianchi N., Conconi A., Gentile C. Tracking the Best Hyperplane with a Simple Budget Perceptron. In Proc. of the 19th Conference on Learning Theory, 483–498p, 2006.

Csat´o L., Opper M. Sparse on-line Gaussian Processes. Neural Comput 2002; 14: 641–668p.

Langford J., Li L., Zhang T. Sparse Online Learning via Truncated Gradient. In Adv Neural Info Process Syst 2008; 21: 905–912p.

Downloads

Published

2015-12-28

Issue

Vol. 1 No. 3 (2014)

Section

Research Articles

License


Declaration and Copyright Transfer Form

(to be completed by authors)

I/ We, the undersigned author(s) of the submitted manuscript, hereby declare, that the above manuscript which is submitted for publication in the STM Journals(s), is not published already in part or whole (except in the form of abstract) in any journal or magazine for private or public circulation, and, is not under consideration of publication elsewhere.

  • I/We will not withdraw the manuscript after 1 week of submission as I have read the Author Guidelines and will adhere to the guidelines.
  • I/We Author(s ) have niether given nor will give this manuscript elsewhere for publishing after submitting in STM Journal(s).
  • I/ We have read the original version of the manuscript and am/ are responsible for the thought contents embodied in it. The work dealt in the manuscript is my/ our own, and my/ our individual contribution to this work is significant enough to qualify for authorship.
  •  I/We also agree to the authorship of the article in the following order:

Author’s name                                    

 

1. ________________

2. ________________

3. ________________

4. ________________

We  Author(s) tick this box and would request you to consider it as our signature as we agree to the terms of this Copyright Notice, which will apply to this submission if and when it is published by this journal.