Risk measurement in Bitcoin market by fusing LSTM with the joint-regression-combined forecasting model

PurposeThe purpose of the paper is to better measure the risks and volatility of the Bitcoin market by using the proposed novel risk measurement model.Design/methodology/approachThe joint regression analysis of value at risk (VaR) and expected shortfall (ES) can effectively overcome the non-elicitability problem of ES to better measure the risks and volatility of financial markets. And because of the incomparable advantages of the long- and short-term memory (LSTM) model in processing non-linear time series, the paper embeds LSTM into the joint regression combined forecasting framework of VaR and ES, constructs a joint regression combined forecasting model based on LSTM for jointly measuring VaR and ES, i.e. the LSTM-joint-combined (LSTM-J-C) model, and uses it to investigate the risks of the Bitcoin market.FindingsEmpirical results show that the proposed LSTM-J-C model can improve forecasting performance of VaR and ES in the Bitcoin market more effectively compared with the historical simulation, the GARCH model and the joint regression combined forecasting model.Social implicationsThe proposed LSTM-J-C model can provide theoretical support and practical guidance to cryptocurrency market investors, policy makers and regulatory agencies for measuring and controlling cryptocurrency market risks.Originality/valueA novel risk measurement model, namely LSTM-J-C model, is proposed to jointly estimate VaR and ES of Bitcoin. On the other hand, the proposed LSTM-J-C model provides risk managers more accurate forecasts of volatility in the Bitcoin market.

Network Security Situation Prediction Model Based on EMD and ELPSO Optimized BiGRU Neural Network

The traditional network security situation prediction method depends on the accuracy of historical situation values, and there are correlations and differences in importance among various network security factors. To solve the above problems, a combined forecasting model based on Empirical Mode Decomposition and improved Particle Swarm Optimization (ELPSO) to optimize BiGRU neural network (EMD-ELPSO-BiGRU) is proposed. Firstly, the model decomposes the network security situation data sequence into a series of intrinsic modal components by empirical mode decomposition; Then, the prediction model of the BiGRU neural network is established for each modal component, and an improved Particle Swarm Optimization Algorithm (ELPSO) is proposed to optimize the super parameters of BiGRU neural network. Finally, the prediction results of each modal component are superimposed to obtain the final prediction value of the network security situation. In the experiment, on the one hand, ELPSO is compared with other particle swarm optimization algorithms, and the results show that ELPSO has better optimization performance; On the other hand, through simulation test and comparison between EMD-ELPSO-BiGRU and other traditional forecasting methods, the results show that the established combined forecasting model has higher forecasting accuracy.

Single Versus Combined Forecasting: The Case Of Wind Run Data

This article compared single to combined forecasts of wind run using artificial neural networks, decomposition, Holt-Winters’ and SARIMA models. The artificial neural networks utilized the feedback framework while decomposition and Holt-Winters’ approaches utilized their multiplicative versions. Holt-Winters’ performed best of single models but ranked fourth, of all fifteen models (single and combined). The combination of decomposition and Holt-Winters’ models ranked best of all two-model combinations and second of all models. Combination of decomposition, Holt-Winters’ and SARIMA performed best of three-model combinations and ranked first, of all models. The only combination of four models ranked third of all models. The accuracy of single forecast should not be underestimated as a single model (Holt-Winters’) outperformed eleven combined models. It is therefore, evident that inclusion of additional model forecast does not necessarily improve combined forecast accuracy. In any modeling situation, single and combined forecasts should be allowed to compete.

Construction and Simulation of Economic Statistics Measurement Model Based on Time Series Analysis and Forecast

Time series follow the basic principles of mathematical statistics and can provide a set of scientifically based dynamic data processing methods. Using this method, various types of data can be approximated by corresponding mathematical models, and then, the internal structure and complex characteristics of the data can be understood essentially, so as to achieve the purpose of predicting its development trend. This paper mainly studies the combined forecasting model based on the time series model and its application. First, the application prospects and research status of the combined forecasting model, the source of time series analysis, and the status of research development at home and abroad are given, and the purpose and significance of the research topic and the research content are summarized. Then, the paper gives the relevant theories about the ARIMA model and the basic principles of model recognition and explains the method of time series smoothing. Finally, the paper uses the ARIMA model to identify and fit the time series data and then the gray forecast model to fit and predict the time series data. Finally, by assigning reasonable weights and combining these methods, a combined forecasting model is proposed and carried out.

A Combined Fuzzy GMDH Neural Network and Grey Wolf Optimization Application for Wind Turbine Power Production Forecasting Considering SCADA Data

A cost-effective and efficient wind energy production trend leads to larger wind turbine generators and drive for more advanced forecast models to increase their accuracy. This paper proposes a combined forecasting model that consists of empirical mode decomposition, fuzzy group method of data handling neural network, and grey wolf optimization algorithm. A combined K-means and identifying density-based local outliers is applied to detect and clean the outliers of the raw supervisory control and data acquisition data in the proposed forecasting model. Moreover, the empirical mode decomposition is employed to decompose signals and pre-processing data. The fuzzy GMDH neural network is a forecaster engine to estimate the future amount of wind turbines energy production, where the grey wolf optimization is used to optimize the fuzzy GMDH neural network parameters in order to achieve a lower forecasting error. Moreover, the model has been applied using actual data from a pilot onshore wind farm in Sweden. The obtained results indicate that the proposed model has a higher accuracy than others in the literature and provides single and combined forecasting models in different time-steps ahead and seasons.

Association of Preoperative NANOG-Positive Circulating Tumor Cell Levels With Recurrence of Hepatocellular Carcinoma

BackgroundCancer stem cells (CSCs) and Circulating tumor cells (CTCs) have been proposed as fundamental causes for the recurrence of hepatocellular carcinoma (HCC). CTCs isolated from patients with HCC illustrate a unique Nanog expression profile analysis. The aim of this study was to enhance the prediction of recurrence and prognosis of the CTC phenotype in patients with HCC by combining Nanog expression into a combined forecasting model.Subjects, Materials, and MethodsWe collected 320 blood samples from 160 patients with HCC cancer before surgery and used CanPatrol™ CTC enrichment technology and in situ hybridization (ISH) to enrich and detect CTCs and CSCs. Nanog expression in all CTCs was also determined. In addition, RT-PCR and immunohistochemistry were used to study the expression of Nanog, E-Cadherin, and N-Cadherin in liver cancer tissues and to conduct clinical correlation studies.ResultsThe numbers of EpCAM mRNA+ CTCs and Nanog mRNA+ CTCs were strongly correlated with postoperative HCC recurrence (CTC number (P = 0.03), the total number of mixed CTCS (P = 0.02), and Nanog> 6.7 (P = 0.001), with Nanog > 6.7 (P = 0.0003, HR = 2.33) being the most crucial marker. There are significant differences in the expression of Nanog on different types of CTC: most Epithelial CTCs do not express Nanog, while most of Mixed CTC and Mesenchymal CTC express Nanog, and their positive rates are 38.7%, 66.7%, and 88.7%, respectively, (P=0.0001). Moreover, both CTC (≤/> 13.3) and Nanog (≤/>6.7) expression were significantly correlated with BCLC stage, vascular invasion, tumor size, and Hbv-DNA (all P < 0.05). In the young group and the old group, patients with higher Nanog expression had a higher recurrence rate. (P < 0.001).ConclusionsThe number of Nanog-positive cells showed positive correlation with the poor prognosis of HCC patients. The detection and analysis of CTC markers (EpCAM and CK8, 18, CD45 Vimentin,Twist and 19) and CSCs markers (NANOG) are of great value in the evaluation of tumor progression.