City lad?s step towards computer aping of human brain

BECAUSE OF the cost, enthusiasm and efforts still required, computer technology may still have some length to go before it can match the human brain's processing power, but earnest efforts continue in this direction world-wide.

Doing his bit to make the dream come true is a young MTech of Allahabad's Indian Institute of Information Technology (IIIT-A), Ankit Mehrotra, who has managed to design a unique neural architecture and algorithm that can make replicating the associate power of the human brain possible in the computers.

Impressed with his work, Ankit has now been invited to participate in 'The 10th World Multi-Conference on Systemics, Cybernetics and Informatics' (WMSCI-2006) all set to kick off in Orlando (Florida), USA, from July 16, 2006. Ankit has also been requested to organise an invited session on his research during the four-day meet and present his paper before IT and computer experts who will gather there from around the world.

"I am simply thrilled. I had submitted my paper on-line to the conference organisers in February 2006 and received the e-mail informing me about the acceptance of the paper in April 2006.

The invitation to present the paper and participate in the conference from Prof Nagib Callaos on behalf of the WMSCI-2006 organising committee arrived on May 4, 2006," informed Ankit, who completed his MTech (Intelligent Systems) from IIIT-A only this year, few hours before catching the Prayagraj Express for New Delhi from where he is to fly to USA on July 11.

"I will present my paper, research for which I did under Prof US Tiwari at IIIT-A, on July 18," added Ankit, a resident of Civil Lines (Allahabad) and whose father Dr Arun Kumar Mehrotra is presently serving as a Joint Director (Health) at Mirzapur.

Explaining his work, Ankit said that his research is an effort to  simulate Human-Brain-Like Learning on to computers. "The human brain learns through building associations.

My effort is to produce a conceptual model for building association between the words and their corresponding meanings over computers. Like brain, the association is in both the directions that is forward as well as backward.
The implementation is done using the Bi-directional Associative Memory (BAM) concept of Neural Networks," he said. Ankit said that BAM provides the architecture for recalling the associated pair and is used for the human brain like associations.

"In Brain the association is bi-directional. For example if laughter signifies happiness, happiness too converges to laughter," he added.