“…The first method is based on old models, such as the autoregressive integrated moving average (ARIMA) (Ilkka and Yli-Olli 1987) and the Cartesian autoregressive integrated moving average search algorithm (CARIMA) (Ostermark 1989). The second method is based on contemporary AI models, such as machine learning models (Parmar et al 2018;Chen et al 2021), artificial neural networks (Vijh et al 2020), deep learning (Jiang 2021;Jing et al 2021), fuzzy logic (Xie et al 2021). Idrees et al (2019), focusing on developing an effective ARIMA model for predicting the volatility of the Indian stock market based on time series data.…”
Section: Related Workmentioning
confidence: 99%
“…This conversion will provide data with the lowest error value when comparing the trained model and test model, if the softmax function is used. The NN output layer comprises the values received after the transformation (Xie et al 2021). If the results obtained are not optimal, the back propagation procedure can be used.…”
Predicting trends in the stock market is becoming complex and uncertain. In response, various artificial intelligence solutions have emerged. A significant solution for predicting the trends of a stock’s volatile and chaotic nature is drawn from deep learning. The present study’s objective is to compare and predict the closing price of the NIFTY 50 index through two significant deep learning methods—long short-term memory (LSTM) and backward elimination LSTM (BE-LSTM)—using 15 years’ worth of per day data obtained from Bloomberg. This study has considered the variables of date, high, open, low, close volume, as well as the 14-period relative strength index (RSI), to predict the closing price. The results of the comparative study show that backward elimination LSTM performs better than the LSTM model for predicting the NIFTY 50 index price for the next 30 days, with an accuracy of 95%. In conclusion, the proposed model has significantly improved the prediction of the NIFTY 50 index price.
“…The first method is based on old models, such as the autoregressive integrated moving average (ARIMA) (Ilkka and Yli-Olli 1987) and the Cartesian autoregressive integrated moving average search algorithm (CARIMA) (Ostermark 1989). The second method is based on contemporary AI models, such as machine learning models (Parmar et al 2018;Chen et al 2021), artificial neural networks (Vijh et al 2020), deep learning (Jiang 2021;Jing et al 2021), fuzzy logic (Xie et al 2021). Idrees et al (2019), focusing on developing an effective ARIMA model for predicting the volatility of the Indian stock market based on time series data.…”
Section: Related Workmentioning
confidence: 99%
“…This conversion will provide data with the lowest error value when comparing the trained model and test model, if the softmax function is used. The NN output layer comprises the values received after the transformation (Xie et al 2021). If the results obtained are not optimal, the back propagation procedure can be used.…”
Predicting trends in the stock market is becoming complex and uncertain. In response, various artificial intelligence solutions have emerged. A significant solution for predicting the trends of a stock’s volatile and chaotic nature is drawn from deep learning. The present study’s objective is to compare and predict the closing price of the NIFTY 50 index through two significant deep learning methods—long short-term memory (LSTM) and backward elimination LSTM (BE-LSTM)—using 15 years’ worth of per day data obtained from Bloomberg. This study has considered the variables of date, high, open, low, close volume, as well as the 14-period relative strength index (RSI), to predict the closing price. The results of the comparative study show that backward elimination LSTM performs better than the LSTM model for predicting the NIFTY 50 index price for the next 30 days, with an accuracy of 95%. In conclusion, the proposed model has significantly improved the prediction of the NIFTY 50 index price.
“…Liu et al [33] presented an application of type two fuzzy neural modelling to predict TAIEX and NASDAQ stock prices based on a given set of training data. Xie et al [34] proposed an approach that integrated a fuzzy neural system with a Hammerstein-Wiener model that formed an indivisible five-layer network, whereby the implication of the fuzzy neural system was realised through a linear dynamic computation of the Hammerstein-Wiener model. The effectiveness of the model was evaluated on three data sets of financial stocks.…”
The article presents the predictive capabilities of a fuzzy multi-criteria evaluation system that operates on the basis of a non-fuzzy neural approach, but also one that is capable of implementing a learning paradigm and working with vague concepts. Within this context, the necessary elements of fuzzy logic are identified and the algebraic formulation of the fuzzy system is presented. It is with the help of the aforementioned that the task of predicting the short-term trend and price of the Tesla share is solved. The functioning of a fuzzy system and fuzzy neural network in the field of time series value prediction is discussed. The authors are inclined to the opinion that, despite the fact that a fuzzy neural network reacts in terms of applicability and effectiveness when solving prediction problems in relation to input data with a faster output than a fuzzy system, and is more “user friendly”, a sufficiently knowledgeable and experienced solver/expert could, by using a fuzzy system, achieve a higher speed of convergence in the learning process than a fuzzy neural network using the minimum range of input data carrying the necessary information. A fuzzy system could therefore be a possible alternative to a fuzzy neural network from the point of view of prediction.
“…They proposed an approach based on a genetic algorithm for adequately conducting the so-called volume-weighted average price trading. Xie et al [20] proposed a method that combines a neuro-fuzzy system with the Hammerstein-Wiener model to create a five-layer network. The proposed model addresses the limitations of conventional neuro-fuzzy systems by realizing their implications through the linear dynamic computation of the Hammerstein-Wiener model.…”
Forecasting stock market prices is an exciting knowledge area for investors and traders. Successful predictions lead to high financial revenues and prevent investors from market risks. This paper proposes a novel hybrid stock prediction model that improves prediction accuracy. The proposed method consists of three main components, a noise-filtering technique, novel features, and machine learning-based prediction. We used a fully modified Hodrick–Prescott filter to smooth the historical stock price data by removing the cyclic component from the time series. We propose several new features for stock price prediction, including the return of firm, return open price, return close price, change in return open price, change in return close price, and volume per total. We investigate traditional and deep machine learning approaches for prediction. Support vector regression, auto-regressive integrated moving averages, and random forests are used for conventional machine learning. Deep learning techniques comprise long short-term memory and gated recurrent units. We performed several experiments with these machine learning algorithms. Our best model achieved a prediction accuracy of 70.88%, a root-mean-square error of 0.04, and an error rate of 0.1.
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.