News @ USP


Three papers for international conference

Back to News »


Dr Rohitash Chandra (left) and Masters Student Shelvin Chand have their papers selected in the International Joint Conference on Neural Networks (IJCNN) to be held in China later this year.

The University of the South Pacific’s academic Dr Rohitash Chandra and Master of Science student Shelvin Chand’s hard work is paying off after their papers got accepted in the International Joint Conference on Neural Networks (IJCNN) to be held in China later this year.

The IJCNN is the premier conference in the area of Artificial Neural Networks and is also ranked as A-tier publication.

“In Computer Science, due to the rapid change in technology, fast publication method is needed.  Conferences in Computer Science have the same peer review process as journals. Full research papers (usually 8-10 pages) are peer reviewed and once accepted, the authors are invited to present at the conference,” Team Leader Dr Chandra said

He said, the papers are published and also indexed in major databases at the conference.

“The published papers are referenced similar to journals and also ranked in a similar way according to measures such as the “H” index.  Institute of Electrical and Electronic Engineers (IEEE) IJCNN is A-ranked and is difficult to get papers accepted for publication,” he said. 

The paper is reviewed in terms of quality of the results, paper presentation, innovation, novelty of the proposed method, presentation and quality of writing.

The three papers include:

  • “Cooperative Coevolution of Feed Forward Neural Networks for Financial Time Series Problem” (paper one) By: Mr Shelvin Chand and Dr Rohitash Chandra.

Cooperative Coevolution is used for training feed forward neural networks for financial time series prediction with an innovative android application for notifying potential investors on future market prices.

  •  “Multi-Objective Cooperative Coevolution of Neural Networks for Time Series Prediction” (paper two) By: Mr Shelvin Chand and Dr Rohitash Chandra.

A novel multi-objective optimization method for time series prediction which uses the time-lag feature of time series prediction to improve accuracy and generalization.

  •  “Competitive Two-Island Cooperative Coevolution for training Elman Recurrent Networks for Time Series Prediction” (paper three) By: Dr Rohitash Chandra

A novel competitive-cooperative method for training recurrent neural networks on time series problems where the performance is improved.

Dr Chandra also gave a message to students.

“It is important to aim for top conferences and journals for the publication of your research. Students should not enrol in Masters by Research just for the sake of getting a degree as research requires deep commitment and seriousness,” he said.

Meanwhile, Shelvin said he will be submitting his thesis in April and having his research validated by a reputed body such as the IEEE only makes him more confident of getting his thesis accepted and approved.

“After submitting my Masters thesis, I hope to immediately start my Ph.D. I give all the credit to my supervisor, Dr Chandra. He has taught me everything I know about research. He spent quite a bit of time training me in the area of Artificial Intelligence as well,” Shelvin said.

He said having supportive parents was always a bonus and that he hopes to inspire younger students to take up research when they finish their undergraduate studies.


This news item was published on 20 Mar 2014 10:21:23 am. For more information or any High-Res Images, please contact us on email communications@usp.ac.fj


News @ USP



General News



Newsletter





Disclaimer & Copyright | Contact Us |
© Copyright 2004 - 2017. All Rights Reserved.
The University of the South Pacific
Laucala Campus, Suva, Fiji
Tel: +679 323 1000