This paper proposes an encoder-decoder based sequence-to-sequence model for Grapheme-to-Phoneme (G2P) conversion in Bangla (Exonym: Bengali). G2P models are key components in speech recognition and speech synthesis systems as they describe how words are pronounced. Traditional, rule-based models do not perform well in unseen contexts. We propose to adopt a neural machine translation (NMT) model to solve the G2P problem. We used gated recurrent units (GRU) recurrent neural network (RNN) to build our model. In contrast to joint-sequence based G2P models, our encoder-decoder based model has the flexibility of not requiring explicit graphemeto-phoneme alignment which are not straight forward to perform. We trained our model on a pronunciation dictionary of (approximately) 135,000 entries and obtained a word error rate (WER) of 12.49% which is a significant improvement from the existing rule-based and machine-learning based Bangla G2P models.
The supply of electricity that exceeds the load requirement results in the occurrence of electrical power losses. To provide the appropriate power supply to these needs, there must be a plan for the provision of electricity by making prediction or estimation of electrical load. Therefore the issue of electrical load forecasting becomes very important in the provision of efficient power. In this study, the author tries to build a model of short-term electrical load prediction using artificial neural network (ANN) with learning algorithm levenberg-marquardt (Trainlm), Bayesian regularization (Trainbr) and scaled conjugate gradient (Trainscg). Scope of research data retrieval is limited electrical load on the work area of Serang City. The results of this study show that the JST prediction of levenberg-marquardt (Trainlm) algorithm is better than the calculated prediction using Bayesian regularization (Trainbr) and scaled conjugate gradient (Trainscg) algorithms. The electric load prediction shows that the average error (Trainlm) is 3.37. Thus, it can be concluded that the electrical load prediction using the levenberg-marquardt (Trainlm) JST algorithm is more accurate than that of the Bayesian regularization (Trainbr) JST algorithm and the scaled conjugate gradient (Trainscg)
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.