Stock market trading has been a subject of interest to investors, academicians, and researchers Analysis of the inherent non-linear characteristics of stock market data is a challenging task A large number of learning algorithms are developed to study market behaviours and enhance the prediction accuracy; they have been optimized using swarm and evolutionary computation such as particle swarm optimization (PSO); its global optimization ability with continuous data has been exploited in financial domains Limitations in the existing approaches and potential future research directions for enhancing PSO-based stock market prediction are discussed This article aims at balancing the economics and computational intelligence aspects; it also analyzes the superiority of PSO for stock portfolio optimization, stock price and trend prediction, and other related stock market aspects along with implications of PSO

A Comprehensive Survey on Portfolio Optimization, Stock Price and Trend Prediction Using Particle Swarm Optimization

In the last few years, deep learning has solved seemingly intractable problems, boosting the hope to find approximate solutions to problems that now are considered unsolvable. Earthquake prediction, the Grail of Seismology, is, in this context of continuous exciting discoveries, an obvious choice for deep learning exploration. We review the entire literature of artificial neural network (ANN) applications for earthquake prediction (77 articles, 1994-2019 period) and find two emerging trends: an increasing interest in this domain, and a complexification of ANN models over time, towards deep learning. Despite apparent positive results observed in this corpus, we demonstrate that simpler models seem to offer similar predictive powers, if not better ones. Due to the structured, tabulated nature of earthquake catalogues, and the limited number of features so far considered, simpler and more transparent machine learning models seem preferable at the present stage of research. Those baseline models follow first physical principles and are consistent with the known empirical laws of Statistical Seismology, which have minimal abilities to predict large earthquakes.

Neural Network Applications in Earthquake Prediction (1994–2019): Meta‐Analytic and Statistical Insights on Their Limitations

Neural Network Applications in Earthquake Prediction (1994-2019): Meta-Analytic Insight on their Limitations

In recent years, new technology developments have been proposed and implemented to support people with hearing impairment and speech loss. It is a severe disability to be unable to speak and communicate. This group of people needs a device to help them use a smartphone in the same way as the rest of the population. This idea has pushed technologists to create new tools to help hearing-impaired and speechless people interact more intelligently and effectively with hearing-impaired and visually impaired people. These tools essentially combine all available new ideas behind speech-to-sign conversion, image processing, gesture extraction and sign-to-speech conversion techniques. Speech-to-sign conversion is processed using template-based recognition with peak modulation. In comparison, many algorithms convert speech-to-sign language, and the template-based peak modulated speech recognition method is superior over other methods, because it considers all components/inputs of speech. Since the highest and lowest peaks are calculated, the accuracy of the method is also high. Converted signs are captured, gesture extraction is performed, and gestures are converted to speech using tri-spectral gesture extraction. The proposed model has a very high accuracy of 98.4% and 98.8% in converting speech to sign and sign to speech, respectively, which is a significant difference from the existing neural network methods. These modules (speech to sign to speech) together compose a Triple-S algorithm that can also be used in wireless communication where hard-of-hearing and speechless people may involve themselves in remote communication as any other people would. The proposed algorithm has been trained and tested with outstanding results with 98.6% accuracy in effective communication through AI, where speech and sign recognition are combined.

Smart communication using tri-spectral sign recognition for hearing-impaired people

Deaf and dumb people use sign language as a tool for communication. As per the 2011 census in India, hearing impaired and speech disabled population is 1,998,692 and 5,072,914 respectively (Disabled persons in India, https://www.mospi.gov.in
 ). Normal hearing people do not learn sign language and thus there is a big communication gap between them and deaf and dumb people. Sign language interpreters can fill this gap but it is a very costly affair to hire them. Indian Sign Language (ISL) consists of signs which are made with two hands while other sign languages like American Sign Language consists of signs made with single hand. This work proposes an automatic and efficient computer vision based system to recognize ISL alphabet which can assist this communication. It can further be used as a module of complete ISL recognition system. Phases in ISL alphabet recognition are image acquisition, preprocessing, segmentation, feature extraction and classification. All 26 ISL alphabet have been considered for testing with average accuracy of 80.76%. Results show that the accuracy of 100% is achieved when similar alphabet {C, L, M, N, R, U,Y} are excluded from testing dataset.

A novel approach for ISL alphabet recognition using Extreme Learning Machine

Real time conversion of sign language to speech and prediction of gestures using Artificial Neural Network

Japan is a country that suffers a lot of earthquakes and disasters because it lies across four major tectonic plates. Subduction zones at the Japanese island curves are geologically complex and create various earthquakes from various sources. Earthquake prediction helps in evacuating areas, which are suspected and could save the lives of people. Artificial neural network is a computing model inspired by biological neurons, which learn from examples and can be able to do predictions. In this paper, we present an artificial neural network with PSO-BP model for the prediction of an earthquake in Japan. In PSO-BP model, particle swarm optimization method is used to optimize the input parameters of backpropagation neural network. Information regarding all major, minor and aftershock earthquake is taken into account for the input of backpropagation neural network. These parameters are taken from Japan seismic catalogue provided by USGS (United States Geological Survey) such as latitude, longitude, magnitude, depth, etc., of earthquake.

Abey Abraham

Papers

Real time conversion of sign language to speech and prediction of gestures using Artificial Neural Network

A Particle Swarm Optimization-Backpropagation (PSO-BP) Model for the Prediction of Earthquake in Japan