Conference Papers
Permanent URI for this collectionhttps://idr.nitk.ac.in/handle/123456789/28506
Browse
9 results
Search Results
Item Stock price movements classification using machine and deep learning techniques-the case study of indian stock market(Springer Verlag service@springer.de, 2019) Naik, N.; Mohan, B.R.Stock price movements forecasting is an important topic for traders and stock analyst. Timely prediction in stock yields can get more profits and returns. The predicting stock price movement on a daily basis is a difficult task due to more ups and down in the financial market. Therefore, there is a need for a more powerful predictive model to predict the stock prices. Most of the existing work is based on machine learning techniques and considered very few technical indicators to predict the stock prices. In this paper, we have extracted 33 technical indicators based on daily stock price such as open, high, low and close price. This paper addresses the two problems, first is the technical indicator feature selection and identification of the relevant technical indicators by using Boruta feature selection technique. The second is an accurate prediction model for stock price movements. To predict stock price movements we have proposed machine learning techniques and deep learning based model. The performance of the deep learning model is better than the machine learning techniques. The experimental results are significant improves the classification accuracy rate by 5% to 6%. National Stock Exchange, India (NSE) stocks are considered for the experiment. © Springer Nature Switzerland AG 2019.Item Study of stock return predictions using recurrent neural networks with LSTM(Springer Verlag service@springer.de, 2019) Naik, N.; Mohan, B.R.Stock price returns forecasting is challenging task for day traders to yield more returns. In the past, most of the literature was focused on machine learning algorithm to predict the stock returns. In this work, the recurrent neural network (RNN) with long short term memory (LSTM) is studied to forecast future stock returns. It has the ability to keep the memory of historical stock returns in order to forecast future stock return output. RNN with LSTM is used to store recent stock information than old related stock information. We have considered a recurrent dropout in RNN layers to avoid overfitting in the model. To accomplish the task we have calculated stock return based on stock closing prices. These stock returns are given as input to the recurrent neural network. The objective function of the prediction model is to minimize the error in the model. To conduct the experiment, data is collected from the National Stock Exchange, India (NSE). The proposed RNN with LSTM model outperforms compared to an feed forward artificial neural network. © Springer Nature Switzerland AG 2019.Item Optimal Feature Selection of Technical Indicator and Stock Prediction Using Machine Learning Technique(Springer Verlag service@springer.de, 2019) Naik, N.; Mohan, B.R.Short-term trading is a difficult task due to fluctuating demand and supply in the stock market. These demands and supply are reflected in stock prices. The stock prices may be predicted using technical indicators. Most of the existing literature considered the limited technical indicators to measure short-term prices. We have considered 33 different combinations of technical indicators to predict the stock prices. The paper has two objectives, first is the technical indicator feature selection and identification of the relevant technical indicators by using Boruta feature selection technique. The second objective is an accurate prediction model for stocks. To predict stock prices we have proposed ANN (Artificial Neural Network) Regression prediction model and model performance is evaluated using metrics is Mean absolute error (MAE) and Root mean square error (RMSE). The experimental results are better than the existing method by decreasing the error rate in the prediction to 12%. We have used the National Stock Exchange, India (NSE) data for the experiment. © 2019, Springer Nature Singapore Pte Ltd.Item GARCH Model Identification for Stock Crises Events(Elsevier B.V., 2020) Naik, N.; Mohan, B.R.; Jha, R.A.The stock market crash is a sudden dramatic decline of stock price due to uncertainty in the stock market. The stock prices are the influence of many factors, such as global trends, local trends, and economic conditions. The identification stock crisis is a challenging task for stock traders and investors. The goal of this paper is to forecast stock crisis events. The experiment is carried out in two steps. First is the least square (LS) method, and the least absolute deviation (LAD) is considered to identify a correlation between mean and median. Based on the correlation between mean and median, the GARCH (General autoregression conditional heteroskedasticity) model proposed to calculate the error distribution in stock returns. To identify the appropriate error distribution, we have varied the degree of t distribution parameters. In the second step, the volatility of stock prices is given as input to the GARCH model to forecast the future crisis events. To carry out the proposed experiment, we have considered Infosys and sbi stock. Experiment results reduce the error in predicting stock crises events. © 2020 The Authors. Published by Elsevier B.V.Item GARCH-Model Identification based on Performance of Information Criteria(Elsevier B.V., 2020) Naik, N.; Mohan, B.R.; Jha, R.A.The stock market prices are volatile due to influence by many factors such as global trends, local trends, and economic conditions. Identification of Generalized autoregressive conditional heteroscedasticity(GARCH) order for stock data is a challenging task due to more fluctuation in stock prices and high variance in data. GARCH is considered to model the conditional volatility of a stock time series. Stock markets data often exhibit volatility clustering. Though many models which belong to autoregressive conditional heteroscedasticity (ARCH) family has proposed, but all the previous studies gave their affirmative consent on the performance of GARCH (1,1), which is considered the standard model, maybe because of the belief held by many researchers that the first lag of conditional variance can capture all the volatility clustering. This can be highly misguiding, especially when the stock market data has high order variance. The focus of this work is to make use of existing, well-known Information Criteria (IC) to identify the stock indices data-generating-process whenever the GARCH effect is present. Akaike Informations Criteria (AIC), Bayesian Information Criteria(BIC), and Hannan-Quinn information(HQ) criteria have used for this experiment. We studied different models with different parameter values and observed the abilities of information criterion in choosing the correct model from a given pool of models. For higher-order GARCH models and high sample sizes, AIC was able to correctly predict the model with high probability, while BIC and HQ performed well for smaller order models. © 2020 The Authors. Published by Elsevier B.V.Item Log Periodic Power Law Fitting on Indian Stock Market(Springer, 2020) Naik, N.; Mohan, B.R.Stock price prediction is one of the challenging tasks for researchers and academics due to frequent changes in stock prices. The stock prices are speculation, and it purely depends on the demand and supply of the market during the trading session. Most of the existing work approach is foresting stock prices using machine learning methods. There has been a limited number of studies on stock crisis identification. Log periodic power law (LPPL) is one of the approaches to identify bubbles in the stock market before crises happened. By looking at existing work, we found that LPPL has not applied in the Indian stock market. In this paper, we have considered LPPL to identify a bubble in the Indian stock market. Due to fluctuation in the market, stock price follows the nonlinearity behavior, hence LPPL is considered to fit the equations. The experiment is carried out R Studio platform. © 2020, Springer Nature Singapore Pte Ltd.Item Time Series Forecasting Using Markov Chain Probability Transition Matrix with Genetic Algorithm Optimisation(Springer Science and Business Media Deutschland GmbH, 2021) Saini, G.; Yadav, N.; Mohan, B.R.; Naik, N.In this paper we are going to discuss the prediction of the financial time series using the Markov chain changing transition matrix model using genetic algorithm. During initial phase of the algorithm we will create the window of fix size with fixed number of state. The basic aim of this paper is to reduce the time taken to find the best window size and best number of states in the window by using the genetic algorithm. This paper produce the approach so that investor can save their time to predict the series without manual activity. To demonstrate the genetic algorithm optimisation we used the historical index data: national stock exchange(NSE50). The Nifty data contained 1239 candles starting from January 1,2015 and ending December 31, 2019. Data was downloaded from [ https://www1.nseindia.com/ ]. In this case we observed the better investment strategy using the first order Markov chain model and reducing the execution time by using the genetic algorithm. © 2021, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.Item Hybrid Model of Multifactor Analysis with RNN-LSTM to Predict Stock Price(Springer Science and Business Media Deutschland GmbH, 2022) Singh, N.; Mohan, B.R.; Naik, N.Prediction on the stock market is one of the most difficult tasks to do in real life. There are so many aspects on which the stock market depends—physical factors versus psychological, rational, and irrational behavior, etc. Proposed research work consists of different aspects on which stock markets are based on. It consists of three models to forecast a stock price on State Bank of India (SBI) stock data. In the current research, we proposed a hybrid model followed by recurrent neural network-long short-term memory (RNN-LSTM) to predict a next-day closing price of SBI. A hybrid model is the combination of two different aspects related to the prediction of stock price. The first technique used other companies’ stock data to predict the target company’s next-day closing price. Other companies lie in the same sector so that they are correlated to each other. For training and testing, we have used multilayer perceptron (MLP) regression model. It is a neural network model in deep learning. The second technique is to predict the stock price of an SBI company using historical data of the target company followed by the auto-regressive integrated moving average—gated recurrent unit (ARIMA-GRU) model. ARIMA-GRU model is a combined model which gives better accuracy for predicting stock price data. In the hybrid model, we take the result of both the models as an input. This paper aims to compare the proposed hybrid model with other two single-aspect models on which stock price depends and proves in terms of accuracy that the hybrid model of all aspects gives better results in comparison to single-aspect models. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.Item Attention-Based Bitemporal Image Deep Feature-Level Change Detection for High Resolution Imagery(Springer Science and Business Media Deutschland GmbH, 2023) Naik, N.; Chandrasekaran, K.; Venkatesan, M.; Prabhavathy, P.To understand the intricacy of changes on the surface of the land, change detection is an important field in the area of remote sensing. Bitemporal remote sensing images are resourceful information to perform the analysis related to classification and change detection. Most of the architectures proposed for improving the performance of change detection in high resolution images pose a challenge due to composite texture features and finer image details. In this paper, we propose a change detection approach for bitemporal images using supervised learning. Firstly, extraction of the features is performed using a pretrained neural network. Then, the extracted features are provided to a (DSDEN) deep supervised-based difference evaluation network. Then, channel and spatial-based attention components are incorporated for fusing the difference image features with the deep features of raw images for the reconstruction of the final change map. The experimental evaluation on public “LEVIR-CD†dataset demonstrates the effectiveness and superiority over traditional methods. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.