Improving Transformer Based End-to-End Code-Switching Speech Recognition Using Language Identification

A Recurrent Neural Networks (RNN) based attention model has been used in code-switching speech recognition (CSSR). However, due to the sequential computation constraint of RNN, there are stronger short-range dependencies and weaker long-range dependencies, which makes it hard to immediately switch languages in CSSR. Firstly, to deal with this problem, we introduce the CTC-Transformer, relying entirely on a self-attention mechanism to draw global dependencies and adopting connectionist temporal classification (CTC) as an auxiliary task for better convergence. Secondly, we proposed two multi-task learning recipes, where a language identification (LID) auxiliary task is learned in addition to the CTC-Transformer automatic speech recognition (ASR) task. Thirdly, we study a decoding strategy to combine the LID into an ASR task. Experiments on the SEAME corpus demonstrate the effects of the proposed methods, achieving a mixed error rate (MER) of 30.95%. It obtains up to 19.35% relative MER reduction compared to the baseline RNN-based CTC-Attention system, and 8.86% relative MER reduction compared to the baseline CTC-Transformer system.

Evolutionary Computation Access on Incremental Map Reduce for Mining Large Scale Data

In recent era, data updates arrive constantly from different areas like social network, finance, healthcare, ecommerce etc… Hence the data becomes large and computation on it becomes difficult. A framework for mining data earlyand to refresh the computed result with the new data arrival is proposed. The framework includes an incremental mapreduce method on hadoop with evolutionary computation algorithm for reduction in time complexity and increased accuracy. Proposed approach is a key pair level incremental iterative processing to Mapreduce for mining big data and uses particle swarm optimization to avoid recomputation from scratch on the new data arrived. Thereby the I/O overhead gets reduced for accessing predefined states. Experimental results were tested on three iterative algorithms in hadoop showed good performance compared to traditional mapreduce with sequential computation access

Verification of internal risk measure estimates

This paper concerns sequential computation of risk measures for financial data and asks how, given a risk measurement procedure, we can tell whether the answers it produces are ‘correct’. We draw the distinction between ‘external’ and ‘internal’ risk measures and concentrate on the latter, where we observe data in real time, make predictions and observe outcomes. It is argued that evaluation of such procedures is best addressed from the point of view of probability forecasting or Dawid’s theory of ‘prequential statistics’ [