For several years the modeling as well as forecasting of the prices of stocks have been extremely challenging for the business community and researchers as a result of the existence of noise in ...samples and also the non-stationary behaviour of information samples. Notwithstanding these drawbacks with improved deep learning, it is now possible to design schemes that will efficiently perform the feature learning task. For this work, we proposed a brand-new end to end algorithm labeled EHTS toward solving the stock price forecasting problem. The AB-CNN and CB-LSTM modules extract features from the stock price dataset and soon after amalgamating the results. Thus, the output of the concatenation stage was feed into the concluding stage which is a stand-alone MLP module. The inclusion of the LSTM and Attention Mechanism in our architecture is to extract long-range and exceptionally long-term stock price information. We experiment the proposed algorithm on two popular stocks both from the NYSE stock market namely “Johnson & Johnson” code-named, “JNJ” and the Bank of America (BAC). In terms of the rMSE, MAE and MAPE error metrics, our proposed scheme gives the lowest error value in all for all datasets. Also, five percentage training window sizes are experimented and EHTS outperforms all the baseline schemes for the different window sizes in all the two datasets with the 70% window size having the highest performance. In terms of number of epochs, EHTS uses the lowest number of epochs for training than the other schemes in all the datasets. Finally, we as well study our stock’s information to point out short-range trading opportunities by performing simulations on our stock price data. The metrics considered in the simulation are as follows: Moving Average (MA), Moving Average Convergence Divergence (MACD) curve, MACD histogram, Signal line, Relative Strength Index (RSI), Returns (R), Annual Returns (AR), Sharpe Ratio (SR), Annual Volatility (V), Maximum DrawDown (MDD) and Daily WinningRate (DWR). For all the aforementioned metrics, EHTS performs better than the baselines. Experimental results revealed that our proposed scheme outperformed the stand-alone deep learning schemes, statistical algorithms, and machine learning models from the beginning to the end.
This study considers the parameter estimation of a multi-variable output-error-like system with autoregressive moving average noise. In order to solve the problem of the information vector containing ...unknown variables, a least squares-based iterative algorithm is proposed by using the iterative search. The original system is divided into several subsystems by using the decomposition technique. However, the subsystems contain the same parameter vector, which poses a challenge for the identification problem, the approach taken here is to use the coupling identification concept to cut down the redundant parameter estimates. In addition, the recursive least squares algorithm is provided for comparison. The simulation results indicate that the proposed algorithms are effective.
Electromechanical modes are inherent to any interconnected power systems which provide a measure of the small-signal stability margin of the system. A number of algorithms have been developed for the ...estimation of these modes using synchrophasor measurements. However, most of these algorithms are not designed to operate in the presence of forced oscillations (FO). These FOs are results of periodic rogue input driving the system. When FOs are present, estimates of system modes can be biased depending on the frequency and the amplitude of the FOs. To tackle this problem, a new algorithm is proposed in this paper to estimate system modes in the presence of FOs. In the proposed method, the over-determined modified Yule-Walker method which is used to estimate autoregressive coefficients of an autoregressive moving average (ARMA) signal model is extended to an ARMA with exogenous input (ARMAX) model that incorporates the presence of FOs. Two versions of the proposed method are included in this paper based on the requirement of the information of the duration of FOs in the signal. Results obtained by implementing the proposed algorithm on simulated data and real-world data validate the effectiveness of both versions of the proposed method.
•A generalized nonlinear model of the random drift is built by a dynamic RNN.•The RNN model is combined with UKF to filter the random drift in real time.•Experiments are carried out to verify the ...effectiveness of the novel algorithm.
The presence of the stochastic errors in MEMS (Micro Electro Mechanical Systems) gyroscopes makes the improvement of the measurement precision challenging. This paper addresses a novel method to estimate and compensate the random drift of MEMS gyroscopes in real time, combining unscented Kalman filter (UKF) with recurrent neural network (RNN). In the proposed method, the random drift is regarded as a generalized nonlinear autoregressive moving average (NARMA) model, and its optimal predictor is realized by a dynamic RNN. To compensate the random drift in real time, the RNN model is brought into the framework of UKF, for establishing the state equation of the improved UKF. The novelty of this paper is that a strategy is presented to guarantee the validity of the combination of UKF and RNN. The effectiveness and superiorities of the proposed method are verified by experiments.
Periodicity and trend are features describing an observed sequence, and extracting these features is an important issue in many scientific fields. However, it is not an easy task for existing methods ...to analyse simultaneously the trend and dynamics of the periodicity such as time varying frequency and amplitude, and the adaptivity of the analysis to such dynamics and robustness to heteroscedastic dependent errors are not guaranteed. These tasks become even more challenging when there are multiple periodic components. We propose a non‐parametric model to describe the dynamics of multicomponent periodicity and investigate the recently developed synchro‐squeezing transform in extracting these features in the presence of a trend and heteroscedastic dependent errors. The identifiability problem of the non‐parametric periodicity model is studied, and the adaptivity and robustness properties of the synchro‐squeezing transform are theoretically justified in both discrete and continuous time settings. Consequently we have a new technique for decoupling the trend, periodicity and heteroscedastic, dependent error process in a general non‐parametric set‐up. Results of a series of simulations are provided, and the incidence time series of varicella and herpes zoster in Taiwan and respiratory signals observed from a sleep study are analysed.
Atmospheric particulate matter (PM) is one of the pollutants that may have a significant impact on human health. Data collected over seven years in a city of the north of Spain is analyzed using four ...different mathematical models: vector autoregressive moving-average (VARMA), autoregressive integrated moving-average (ARIMA), multilayer perceptron (MLP) neural networks and support vector machines (SVMs) with regression. Measured monthly average pollutants and PM10 (particles with a diameter less than 10μm) concentration are used as input to forecast the monthly averaged concentration of PM10 from one to seven months ahead. Simulations showed that the SVM model performs better than the other models when forecasting one month ahead and also for the following seven months.
Display omitted
•Four models based on SVM, MLP, VARMA and ARIMA are built for forecasting of the PM10 concentration in the city of Oviedo.•PM10 have impacts on climate and precipitation that adversely affect human health.•The description of the air quality is of real interest for the effective safety management of the air pollution in cities.•The results show that the SVM model was better than the other models to forecast PM10 concentration.
The application of graph convolutional networks (GCNs) to hyperspectral image (HSI) classification is a heavily researched topic. However, GCNs are based on spectral filters, which are ...computationally costly and fail to suppress noise effectively. In addition, the current GCN-based methods are prone to oversmoothing (the representation of each node tends to be congruent) problems. To circumvent these problems, a novel semi-supervised locality-preserving dense graph neural network (GNN) with autoregressive moving average (ARMA) filters and context-aware learning (DARMA-CAL) is proposed for HSI classification. In this work, we introduce the ARMA filter instead of a spectral filter to apply to GNNs. The ARMA filter can better capture the global graph structure and is more robust to noise. More importantly, the ARMA filter can simplify calculations compared with the spectral filter. In addition, we show that the ARMA filter can be approximated by a recursive method. Furthermore, we propose a dense structure, which not only implements the ARMA filter in the structure, but is also locality-preserving. Finally, we design a layerwise context-aware learning mechanism to extract the useful local information generated by each layer of the dense ARMA network. The experimental results on three real HSI datasets show that DARMA-CAL outperforms the compared state-of-the-art methods.
Several machine learning and deep learning models were reported in the literature to forecast COVID-19 but there is no comprehensive report on the comparison between statistical models and deep ...learning models. The present work reports a comparative time-series analysis of deep learning techniques (Recurrent Neural Networks with GRU and LSTM cells) and statistical techniques (ARIMA and SARIMA) to forecast the country-wise cumulative confirmed, recovered, and deaths. The Gated Recurrent Units (GRU), Long Short-Term Memory (LSTM) cells based on Recurrent Neural Networks (RNN), ARIMA and SARIMA models were trained, tested, and optimized to forecast the trends of the COVID-19. We deployed python to optimize the parameters of ARIMA which include (p, d, q) representing autoregressive and moving average terms and parameters of SARIMA model include additional seasonal terms which are denoted by (P, D, Q). Similarly, for LSTM and GRU based RNN models’ parameters (number of layers, hidden size, learning rate and number of epochs) are optimized by deploying PyTorch machine learning framework. The best model was chosen based on the lowest Mean Square Error (MSE) and Root Mean Squared Error (RMSE) values. For most of the time-series data of the countries, deep learning-based models LSTM and GRU outperformed statistical ARIMA and SARIMA models, with an RMSE values that are 40 folds less than that of the ARIMA models. But for some countries statistical (ARIMA, SARIMA) models outperformed deep learning models. Further, we emphasize the importance of various factors such as age, preventive measures and healthcare facilities etc. that play vital role on the rapid spread of COVID-19 pandemic.
The mixed control chart is proposed to improve detection performance with fewer process shifts. In this study, we proposed the modified exponentially weighted moving average - moving average control ...chart (MMEM), a new mixed control chart for observing the changes in the process mean. Average run length, standard deviation of run length, and median run length can be used to examine the effectiveness of detecting changes in the proposed chart with Shewhart, Moving Average (MA), Modified Exponentially Weighted Moving Average (MEWMA), and Mixed Moving Average - Modified Exponentially Weighted Moving Average (MMME) control charts in parametric and nonparametric distributions that use Monte Carlo simulation. The results demonstrate that the proposed chart outperforms other control charts mostly in the detection of small-to-moderate shifts. To illustrate the application of the proposed chart, chemical process temperature data and dataset on survival times of a group of patients suffering from head and neck cancer disease and treated with radiotherapy were provided, and it was discovered that the proposed chart performs better than other control charts.
This paper is concerned by the modal identification of time-varying mechanical systems. Based on previous works about autoregressive moving average models in vector form (ARMAV) for the modal ...identification of linear time invariant systems, and time-varying autoregressive moving average models (TV-ARMA) for the identification of nonstationary systems, a time-varying ARMAV (TV-ARMAV) model is presented for the multivariate identification of time-varying systems. It results in the identification of not only the time-varying poles of the system but also of its respective time-varying mode shapes. The method is applied on a time-varying structure composed of a beam on which a mass is moving.
•The modal identification of time-varying systems is addressed.•Multivariate autoregressive moving-average models are used for the identification.•The time-varying behavior is handled by the use of basis functions.•All the time-varying modal parameters are identified (poles and mode shapes).•The method is successfully tested on an experimental structure.