Implementation of Extreme Learning Machine and Software as a Service for Recognition of Tamil Handwritten words

The conventional cloud computing paradigm has spurred the development of innovative architectures for next-generation cloud computing. These novel designs can effectively manage vast amounts of data, which was previously beyond the capabilities of conventional architectures. To tackle this challenge, feed forward neural network algorithms have emerged as a cutting-edge solution for processing massive data sets. As a result, they are now widely employed for solving various tasks, particularly in the pitch of recognition. One promising research application in the domain of postal automation is the recognition of handwritten town names. To address this challenge, we have employed an Extreme Learning Machine to enhance the concert of feed forward neural networks. Our approach has yielded superior results compared to state-of-the-art algorithms.