A Survey of Forex and Stock Price Prediction Using Deep Learning

Predictions of stock and foreign exchange (Forex) have always been a hot and profitable area of study. Deep learning applications have been proven to yield better accuracy and return in the field of financial prediction and forecasting. In this survey, we selected papers from the Digital Bibliography & Library Project (DBLP) database for comparison and analysis. We classified papers according to different deep learning methods, which included Convolutional neural network (CNN); Long Short-Term Memory (LSTM); Deep neural network (DNN); Recurrent Neural Network (RNN); Reinforcement Learning; and other deep learning methods such as Hybrid Attention Networks (HAN), self-paced learning mechanism (NLP), and Wavenet. Furthermore, this paper reviews the dataset, variable, model, and results of each article. The survey used presents the results through the most used performance metrics: Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), Mean Absolute Error (MAE), Mean Square Error (MSE), accuracy, Sharpe ratio, and return rate. We identified that recent models combining LSTM with other methods, for example, DNN, are widely researched. Reinforcement learning and other deep learning methods yielded great returns and performances. We conclude that, in recent years, the trend of using deep-learning-based methods for financial modeling is rising exponentially.

Download Full-text

PP15 Predicting variations of calls to an ambulance service in the UK caused by circulating infections using-deep learning methods

Emergency Medicine Journal ◽

10.1136/emermed-2019-999abs.15 ◽

2019 ◽

Vol 36 (10) ◽

pp. e7.2-e7

Author(s):

Thilo Reich ◽

Marcin Budka

Keyword(s):

Deep Learning ◽

Short Term Memory ◽

Prehospital Care ◽

Machine Learning Algorithms ◽

Ambulance Service ◽

Percentage Error ◽

Learning Methods ◽

Average Percentage ◽

Average Percentage Error ◽

The Uk

BackgroundDigital patient records in the ambulance service have opened up new opportunities for prehospital care. Previously it was demonstrated that prehospital pyrexia numbers are linked to an increase in overall calls to the ambulance service. This study aims to predict the future number of calls using deep-learning methods.MethodsTemperature readings for 280,447 patients were generously provided by the South Western Ambulance Service Trust. The data covered the time between 05/01/2016 and 30/04/2017 with overall 44,472 patients being pyretic. A rolling window of 10 days was applied to daily sums for both pyretic and apyretic patients. These windows were used as input features to train machine-learning algorithms predicting the number of calls 10 days ahead. Algorithms tested include Linear Regression (LR), basic Recurrent Neural Networks (RNN), Long Short Term Memory (LSTM) and Gated Recurrent Unit (GRU) architectures. A genetic approach was used to optimise the architecture, in which parameters were randomly modified and over several generations the best performing algorithm will be selected to be further manipulated. To assess performance the Mean Average Percentage Error (MAPE) was used.ResultsThe initial analysis showed that the total patient number and pyretic patient numbers are correlated. The best performing algorithms with varying numbers of hidden units had the following MAPE in comparison to simple LR: LR=19.4%, LSTM (104 units) = 6.1%, RNN (79 units)=6.01%, GRU (80 units)=5.97%.ConclusionsThese preliminary results suggest that deep-learning methods allow to predict the variations in total number of calls caused by circulating infections. Further investigations will aim to confirm these findings. Once fully verified these algorithms could play a major role in operational planning of any ambulance service by predicting increases in demand.

Download Full-text

A Comparative Study of Bitcoin Price Prediction Using Deep Learning

Mathematics ◽

10.3390/math7100898 ◽

2019 ◽

Vol 7 (10) ◽

pp. 898 ◽

Cited By ~ 12

Author(s):

Suhwan Ji ◽

Jongmin Kim ◽

Hyeonseung Im

Keyword(s):

Neural Network ◽

Deep Learning ◽

Prediction Models ◽

Short Term Memory ◽

Learning Methods ◽

The Public ◽

Price Prediction ◽

Profitability Analysis ◽

The Media ◽

Long Short Term Memory

Bitcoin has recently received a lot of attention from the media and the public due to its recent price surge and crash. Correspondingly, many researchers have investigated various factors that affect the Bitcoin price and the patterns behind its fluctuations, in particular, using various machine learning methods. In this paper, we study and compare various state-of-the-art deep learning methods such as a deep neural network (DNN), a long short-term memory (LSTM) model, a convolutional neural network, a deep residual network, and their combinations for Bitcoin price prediction. Experimental results showed that although LSTM-based prediction models slightly outperformed the other prediction models for Bitcoin price prediction (regression), DNN-based models performed the best for price ups and downs prediction (classification). In addition, a simple profitability analysis showed that classification models were more effective than regression models for algorithmic trading. Overall, the performances of the proposed deep learning-based prediction models were comparable.

Download Full-text

sEMG-Based Neural Network Prediction Model Selection of Gesture Fatigue and Dataset Optimization

Computational Intelligence and Neuroscience ◽

10.1155/2020/8853314 ◽

2020 ◽

Vol 2020 ◽

pp. 1-17

Author(s):

Fujun Ma ◽

Fanghao Song ◽

Yan Liu ◽

Jiahui Niu

Keyword(s):

Neural Network ◽

Energy Consumption ◽

Prediction Model ◽

Mean Square Error ◽

Short Term Memory ◽

Experimental Results ◽

Support Vector ◽

Mean Square ◽

Semg Signals ◽

Research Studies

The fatigue energy consumption of independent gestures can be obtained by calculating the power spectrum of surface electromyography (sEMG) signals. The existing research studies focus on the fatigue of independent gestures, while the research studies on integrated gestures are few. However, the actual gesture operation mode is usually integrated by multiple independent gestures, so the fatigue degree of integrated gestures can be predicted by training neural network of independent gestures. Three natural gestures including browsing information, playing games, and typing are divided into nine independent gestures in this paper, and the predicted model is established and trained by calculating the energy consumption of independent gestures. The artificial neural networks (ANNs) including backpropagation (BP) neural network, recurrent neural network (RNN), and long short-term memory (LSTM) are used to predict the fatigue of gesture. The support vector machine (SVM) is used to assist verification. Mean square error (MSE), root mean square error (RMSE), and mean absolute error (MAE) are utilized to evaluate the optimal prediction model. Furthermore, the different datasets of the processed sEMG signal and its decomposed wavelet coefficients are trained, respectively, and the changes of error functions of them are compared. The experimental results show that LSTM model is more suitable for gesture fatigue prediction. The processed sEMG signals are appropriate for using as the training set the fatigue degree of one-handed gesture. It is better to use wavelet decomposition coefficients as datasets to predict the high-dimensional sEMG signals of two-handed gestures. The experimental results can be applied to predict the fatigue degree of complex human-machine interactive gestures, help to avoid unreasonable gestures, and improve the user’s interactive experience.

Download Full-text

Dissolved Oxygen Forecasting in Aquaculture: A Hybrid Model Approach

Applied Sciences ◽

10.3390/app10207079 ◽

2020 ◽

Vol 10 (20) ◽

pp. 7079

Author(s):

Elias Eze ◽

Tahmina Ajmal

Keyword(s):

Dissolved Oxygen ◽

Mean Square Error ◽

Short Term Memory ◽

Sensor Data ◽

Percentage Error ◽

Forecasting Model ◽

Mean Square ◽

Intrinsic Mode Functions ◽

Short Term ◽

Do Concentration

Dissolved oxygen (DO) concentration is a vital parameter that indicates water quality. We present here DO short term forecasting using time series analysis on data collected from an aquaculture pond. This can provide the basis of data support for an early warning system, for an improved management of the aquaculture farm. The conventional forecasting approaches are commonly characterized by low accuracy and poor generalization problems. In this article, we present a novel hybrid DO concentration forecasting method with ensemble empirical mode decomposition (EEMD)-based LSTM (long short-term memory) neural network (NN). With this method, first, the sensor data integrity is improved through linear interpolation and moving average filtering methods of data preprocessing. Next, the EEMD algorithm is applied to decompose the original sensor data into multiple intrinsic mode functions (IMFs). Finally, the feature selection is used to carefully select IMFs that strongly correlate with the original sensor data, and integrate into both inputs for the NN. The hybrid EEMD-based LSTM forecasting model is then constructed. The performance of this proposed model in training and validation sets was compared with the observed real sensor data. To obtain the exact evaluation accuracy of the forecasted results of the hybrid EEMD-based LSTM forecasting model, four statistical performance indices were adopted: mean absolute error (MAE), mean square error (MSE), root mean square error (RMSE), and mean absolute percentage error (MAPE). Results are presented for the short term (12-h) and the long term (1-month) that are encouraging, indicating suitability of this technique for forecasting DO values.

Download Full-text

A Statistics and Deep Learning Hybrid Method for Multivariate Time Series Forecasting and Mortality Modeling

Forecasting ◽

10.3390/forecast4010001 ◽

2021 ◽

Vol 4 (1) ◽

pp. 1-25

Author(s):

Thabang Mathonsi ◽

Terence L. van Zyl

Keyword(s):

Neural Network ◽

Deep Learning ◽

Recurrent Neural Network ◽

Short Term Memory ◽

Hybrid Methods ◽

Prediction Interval ◽

Hybrid Approach ◽

Computational Cost ◽

Exponential Smoothing ◽

Learning Methods

Hybrid methods have been shown to outperform pure statistical and pure deep learning methods at forecasting tasks and quantifying the associated uncertainty with those forecasts (prediction intervals). One example is Exponential Smoothing Recurrent Neural Network (ES-RNN), a hybrid between a statistical forecasting model and a recurrent neural network variant. ES-RNN achieves a 9.4% improvement in absolute error in the Makridakis-4 Forecasting Competition. This improvement and similar outperformance from other hybrid models have primarily been demonstrated only on univariate datasets. Difficulties with applying hybrid forecast methods to multivariate data include (i) the high computational cost involved in hyperparameter tuning for models that are not parsimonious, (ii) challenges associated with auto-correlation inherent in the data, as well as (iii) complex dependency (cross-correlation) between the covariates that may be hard to capture. This paper presents Multivariate Exponential Smoothing Long Short Term Memory (MES-LSTM), a generalized multivariate extension to ES-RNN, that overcomes these challenges. MES-LSTM utilizes a vectorized implementation. We test MES-LSTM on several aggregated coronavirus disease of 2019 (COVID-19) morbidity datasets and find our hybrid approach shows consistent, significant improvement over pure statistical and deep learning methods at forecast accuracy and prediction interval construction.

Download Full-text

Comparing Four Types Methods for Karst NDVI Prediction Based on Machine Learning

Atmosphere ◽

10.3390/atmos12101341 ◽

2021 ◽

Vol 12 (10) ◽

pp. 1341

Author(s):

Yuju Ma ◽

Liyuan Zuo ◽

Jiangbo Gao ◽

Qiang Liu ◽

Lulu Liu

Keyword(s):

Neural Network ◽

Mean Square Error ◽

Human Activities ◽

Vegetation Index ◽

Terrestrial Ecosystem ◽

Coefficient Of Determination ◽

Percentage Error ◽

Support Vector ◽

Mean Square ◽

Response Factors

As a link for energy transfer between the land and atmosphere in the terrestrial ecosystem, karst vegetation plays an important role. Karst vegetation is not only affected by environmental factors but also by intense human activities. The nonlinear characteristics of vegetation growth are induced by the interaction mechanism of these factors. Previous studies of this relationship were not comprehensive, and it is necessary to further explore it using a suitable method. In this study, we selected climate, human activities, topography, and soil texture as the response factors; a nonlinear relationship model between the karst normalized difference vegetation index (NDVI) and these factors was established by applying a back propagation neural network (BPNN), a radial basis function neural network (RBFNN), the random forest (RF) algorithm, and support vector regression (SVR); and then, the karst NDVI was predicted. The coefficient of determination (R2), mean square error (MSE), root mean square error (RMSE), and mean absolute percentage error (MAPE) of the obtained results were calculated, and the mean R2 values of the BPNN, RBFNN, RF, and SVR models were determined to be 0.77, 0.86, 0.89, and 0.91, respectively. Compared with the BPNN, RBFNN, and RF models, the SVR model had the lowest errors, with mean MSE, RMSE, and MAPE values of 0.001, 0.02, and 2.77, respectively. The results show that the BPNN, RBFNN, RF, and SVR models are within acceptable ranges for karst NDVI prediction, but the overall performance of the SVR model is the best, and it is more suitable for karst vegetation prediction.

Download Full-text

Application of deep learning methods to predict ionosphere parameters in real time

E3S Web of Conferences ◽

10.1051/e3sconf/202019602007 ◽

2020 ◽

Vol 196 ◽

pp. 02007

Author(s):

Vladimir Mochalov ◽

Anastasia Mochalova

Keyword(s):

Neural Network ◽

Deep Learning ◽

Real Time ◽

Network Architecture ◽

Short Term Memory ◽

Neural Network Architecture ◽

Short Term ◽

Learning Methods ◽

Term Memory ◽

Long Short Term Memory

In this paper, the previously obtained results on recognition of ionograms using deep learning are expanded to predict the parameters of the ionosphere. After the ionospheric parameters have been identified on the ionogram using deep learning in real time, we can predict the parameters for some time ahead on the basis of the new data obtained Examples of predicting the ionosphere parameters using an artificial recurrent neural network architecture long short-term memory are given. The place of the block for predicting the parameters of the ionosphere in the system for analyzing ionospheric data using deep learning methods is shown.

Download Full-text

A Spectral-Spatial Cascaded 3D Convolutional Neural Network with a Convolutional Long Short-Term Memory Network for Hyperspectral Image Classification

Remote Sensing ◽

10.3390/rs11202363 ◽

2019 ◽

Vol 11 (20) ◽

pp. 2363 ◽

Cited By ~ 2

Author(s):

Wenchao Qi ◽

Xia Zhang ◽

Nan Wang ◽

Mao Zhang ◽

Yi Cen

Keyword(s):

Neural Network ◽

Deep Learning ◽

Convolutional Neural Network ◽

Short Term Memory ◽

Hyperspectral Image ◽

Short Term ◽

Dynamic Learning ◽

Learning Methods ◽

Term Memory ◽

Long Short Term Memory

Deep learning methods used for hyperspectral image (HSI) classification often achieve greater accuracy than traditional algorithms but require large numbers of training epochs. To simplify model structures and reduce their training epochs, an end-to-end deep learning framework incorporating a spectral-spatial cascaded 3D convolutional neural network (CNN) with a convolutional long short-term memory (CLSTM) network, called SSCC, is proposed herein for HSI classification. The SSCC framework employs cascaded 3D CNN to learn the spectral-spatial features of HSIs and uses the CLSTM network to extract sequence features. Residual connections are used in SSCC to accelerate model convergence, with the outputs of previous convolutional layers concatenated as inputs for subsequent layers. Moreover, the data augmentation, parametric rectified linear unit, dynamic learning rate, batch normalization, and regularization (including dropout and L2) methods are used to increase classification accuracy and prevent overfitting. These attributes allow the SSCC framework to achieve good performance for HSI classification within 20 epochs. Three well-known datasets including Indiana Pines, University of Pavia, and Pavia Center were employed to evaluate the classification performance of the proposed algorithm. The GF-5 dataset of Anxin County, obtained from China’s recently launched spaceborne Advanced Hyperspectral Imager, was also used for classification experiments. The experimental results demonstrate that the proposed SSCC framework achieves state-of-the-art performance with better training efficiency than other deep learning methods.

Download Full-text

Forecasting PM2.5 Concentration Using a Single-Dense Layer BiLSTM Method

Electronics ◽

10.3390/electronics10151808 ◽

2021 ◽

Vol 10 (15) ◽

pp. 1808

Author(s):

Aji Teguh Prihatno ◽

Himawan Nurcahyanto ◽

Md. Faisal Ahmed ◽

Md. Habibur Rahman ◽

Md. Morshed Alam ◽

...

Keyword(s):

Mean Square Error ◽

Time Series Data ◽

Short Term Memory ◽

Series Data ◽

Percentage Error ◽

Dense Layer ◽

Mean Square ◽

Time Data ◽

Industrial Grade ◽

Pm2.5 Concentration

In recent times, particulate matter (PM2.5) is one of the most critical air quality contaminants, and the rise of its concentration will intensify the hazard of cleanrooms. The forecasting of the concentration of PM2.5 has great importance to improve the safety of the highly pollutant-sensitive electronic circuits in the factories, especially inside semiconductor industries. In this paper, a Single-Dense Layer Bidirectional Long Short-term Memory (BiLSTM) model is developed to forecast the PM2.5 concentrations in the indoor environment by using the time series data. The real-time data samples of PM2.5 concentrations were obtained by using an industrial-grade sensor based on edge computing. The proposed model provided the best results comparing with the other existing models in terms of mean absolute error, mean square error, root mean square error, and mean absolute percentage error. These results show that the low error of forecasting PM2.5 concentration in a cleanroom in a semiconductor factory using the proposed Single-Dense Layer BiLSTM method is considerably high.

Download Full-text

COVID-19 in Iran: Forecasting Pandemic Using Deep Learning

Computational and Mathematical Methods in Medicine ◽

10.1155/2021/6927985 ◽

2021 ◽

Vol 2021 ◽

pp. 1-16

Author(s):

Rahele Kafieh ◽

Roya Arian ◽

Narges Saeedizadeh ◽

Zahra Amini ◽

Nasim Dadashi Serej ◽

...

Keyword(s):

Deep Learning ◽

Short Term Memory ◽

Percentage Error ◽

Mean Square ◽

Short Term ◽

Average Percentage ◽

The Usa ◽

Long Short Term Memory ◽

Average Percentage Error ◽

Almost All

COVID-19 has led to a pandemic, affecting almost all countries in a few months. In this work, we applied selected deep learning models including multilayer perceptron, random forest, and different versions of long short-term memory (LSTM), using three data sources to train the models, including COVID-19 occurrences, basic information like coded country names, and detailed information like population, and area of different countries. The main goal is to forecast the outbreak in nine countries (Iran, Germany, Italy, Japan, Korea, Switzerland, Spain, China, and the USA). The performances of the models are measured using four metrics, including mean average percentage error (MAPE), root mean square error (RMSE), normalized RMSE (NRMSE), and R 2 . The best performance was found for a modified version of LSTM, called M-LSTM (winner model), to forecast the future trajectory of the pandemic in the mentioned countries. For this purpose, we collected the data from January 22 till July 30, 2020, for training, and from 1 August 2020 to 31 August 2020, for the testing phase. Through experimental results, the winner model achieved reasonably accurate predictions (MAPE, RMSE, NRMSE, and R 2 are 0.509, 458.12, 0.001624, and 0.99997, respectively). Furthermore, we stopped the training of the model on some dates related to main country actions to investigate the effect of country actions on predictions by the model.

Download Full-text