Deep Learning for Financial Time Series Prediction:A State-of-the-Art Review of Standalone and HybridModels

Financial time series prediction,whether for classification or regression,has been a heated research topic over the last decade.While traditional machine learning algorithms have experienced mediocre results,deep learning has largely contributed to the elevation of the prediction performance.Currently,the most up-to-date review of advanced machine learning techniques for financial time series prediction is still lacking,making it challenging for finance domain experts and relevant practitioners to determine which model potentially performs better,what techniques and components are involved,and how themodel can be designed and implemented.This review article provides an overview of techniques,components and frameworks for financial time series prediction,with an emphasis on state-of-the-art deep learning models in the literature from2015 to 2023,including standalonemodels like convolutional neural networks(CNN)that are capable of extracting spatial dependencies within data,and long short-term memory(LSTM)that is designed for handling temporal dependencies;and hybrid models integrating CNN,LSTM,attention mechanism(AM)and other techniques.For illustration and comparison purposes,models proposed in recent studies are mapped to relevant elements of a generalized framework comprised of input,output,feature extraction,prediction,and related processes.Among the state-of-the-artmodels,hybrid models like CNNLSTMand CNN-LSTM-AM in general have been reported superior in performance to stand-alone models like the CNN-only model.Some remaining challenges have been discussed,including non-friendliness for finance domain experts,delayed prediction,domain knowledge negligence,lack of standards,and inability of real-time and highfrequency predictions.The principal contributions of this paper are to provide a one-stop guide for both academia and industry to review,compare and summarize technologies and recent advances in this area,to facilitate smooth and informed implementation,and to highlight future research directions.

Hurst Exponent Analysis of Financial Time Series

Statistical properties of stock market time series and the implication of their Hurst exponents are discussed. Hurst exponents of DJIA (Dow Jones Industrial Average) components are tested using re scaled range analysis. In addition to the original stock return series, the linear prediction errors of the daily returns are also tested. Numerical results show that the Hurst exponent analysis can provide some information about the statistical properties of the financial time series.

Performance evaluation of series and parallel strategies for financial time series forecasting

Background:Improving financial time series forecasting is one of the most challenging and vital issues facing numerous financial analysts and decision makers.Given its direct impact on related decisions,various attempts have been made to achieve more accurate and reliable forecasting results,of which the combining of individual models remains a widely applied approach.In general,individual models are combined under two main strategies:series and parallel.While it has been proven that these strategies can improve overall forecasting accuracy,the literature on time series forecasting remains vague on the choice of an appropriate strategy to generate a more accurate hybrid model.Methods:Therefore,this study’s key aim is to evaluate the performance of series and parallel strategies to determine a more accurate one.Results:Accordingly,the predictive capabilities of five hybrid models are constructed on the basis of series and parallel strategies compared with each other and with their base models to forecast stock price.To do so,autoregressive integrated moving average(ARIMA)and multilayer perceptrons(MLPs)are used to construct two series hybrid models,ARIMA-MLP and MLP-ARIMA,and three parallel hybrid models,simple average,linear regression,and genetic algorithm models.Conclusion:The empirical forecasting results for two benchmark datasets,that is,the closing of the Shenzhen Integrated Index(SZII)and that of Standard and Poor’s 500(S&P 500),indicate that although all hybrid models perform better than at least one of their individual components,the series combination strategy produces more accurate hybrid models for financial time series forecasting.

Prediction of metal futures price volatility and empirical analysis based on symbolic time series of high-frequency

The metal futures price fluctuation prediction model was constructed based on symbolic high-frequency time series using high-frequency data on the Shanghai Copper Futures Exchange from July 2014 to September 2018,and the sample was divided into 194 histogram time series employing symbolic time series.The next cycle was then predicted using the K-NN algorithm and exponential smoothing,respectively.The results show that the trend of the histogram of the copper futures earnings prediction is gentler than that of the actual histogram,the overall situation of the prediction results is better,and the overall fluctuation of the one-week earnings of the copper futures predicted and the actual volatility are largely the same.This shows that the results predicted by the K-NN algorithm are more accurate than those predicted by the exponential smoothing method.Based on the predicted one-week price fluctuations of copper futures,regulators and investors in China’s copper futures market can timely adjust their regulatory policies and investment strategies to control risks.

Modelling and Analysis on Noisy Financial Time Series

Building the prediction model(s) from the historical time series has attracted many researchers in last few decades. For example, the traders of hedge funds and experts in agriculture are demanding the precise models to make the prediction of the possible trends and cycles. Even though many statistical or machine learning (ML) models have been proposed, however, there are no universal solutions available to resolve such particular problem. In this paper, the powerful forward-backward non-linear filter and wavelet-based denoising method are introduced to remove the high level of noise embedded in financial time series. With the filtered time series, the statistical model known as autoregression is utilized to model the historical times aeries and make the prediction. The proposed models and approaches have been evaluated using the sample time series, and the experimental results have proved that the proposed approaches are able to make the precise prediction very efficiently and effectively.

Short-Term Financial Time Series Forecasting Integrating Principal Component Analysis and Independent Component Analysis with Support Vector Regression

Financial time series forecasting could be beneficial for individual as well as institutional investors. But, the high noise and complexity residing in the financial data make this job extremely challenging. Over the years, many researchers have used support vector regression (SVR) quite successfully to conquer this challenge. In this paper, an SVR based forecasting model is proposed which first uses the principal component analysis (PCA) to extract the low-dimensional and efficient feature information, and then uses the independent component analysis (ICA) to preprocess the extracted features to nullify the influence of noise in the features. Experiments were carried out based on 16 years’ historical data of three prominent stocks from three different sectors listed in Dhaka Stock Exchange (DSE), Bangladesh. The predictions were made for 1 to 4 days in advance targeting the short term prediction. For comparison, the integration of PCA with SVR (PCA-SVR), ICA with SVR (ICA-SVR) and single SVR approaches were applied to evaluate the prediction accuracy of the proposed approach. Experimental results show that the proposed model (PCA-ICA-SVR) outperforms the PCA-SVR, ICA-SVR and single SVR methods.

Forecasting High-Frequency Long Memory Series with Long Periods Using the SARFIMA Model

This paper evaluates the efficiency of the SARFIMA model at forecasting high-frequency long memory series with especially long periods. Three other models, the ARFIMA, ARMA and PAR models, are also included to compare their forecasting performances with that of the SARFIMA model. For the artificial SARFIMA series, if the correct parameters are used for estimating and forecasting, the model performs as well as the other three models. However, if the parameters obtained by the WHI estimation are used, the performance of the SARFIMA model falls far behind that of the other models. For the empirical intraday volume series, the SARFIMA model produces the worst performance of all of the models, and the ARFIMA model performs best. The ARMA and PAR models perform very well both for the artificial series and for the intraday volume series. This result indicates that short memory models are competent in forecasting periodic long memory series.

Time Optimal Control in Time Series Movement

The paper analyses time series that exhibit equilibrium states. It analyses the formation of equilibrium and how the system can return to the aforementioned equilibrium. The tool that is used in the aforementioned analysis is time optimal control in the phase plane. It is proved that equilibrium state is sustainable if initial state is not too far from the equilibrium as well as control vector is large enough. On the other hand, if initial state is one standard deviation away from equilibrium state, it is proved that equilibrium cannot be reached. It is the same case with control vector. If it is unbounded, time optimal control cannot be applied. The approach that is introduced represents unconventional method of analysing equilibrium in time series.

基于BSFinformer模型的金融数据特征选择及预测

金融领域的长时间序列预测正在面对复杂的市场和众多金融产品的挑战,传统的时序数据预测方法在处理线性分布数据时表现良好,但对于特征参数冗余和非线性长序列金融产品数据的预测效果有限.为了解决这一问题,提出一种长时间序列预测方法BSFinformer(Boruta-SHAP+Finformer),利用金融数据的时间相关性并综合运用BorutaSHAP,Finformer等技术来完成特征选择及预测功能.该方法首先引入Boruta-SHAP模块,利用XgBoost和SHAP分析方法进行特征选择,从给定的特征集中识别出与金融时间序列预测任务相关的重要特征,并解释这些特征对预测的影响.其次,利用Transformer结构和自注意力机制,改进为Finformer模块,将长序列金融数据分解为趋势、周期和残差成分,结合稀疏自注意力机制.在多个真实金融数据集上进行了实验评估.实验结果显示,BSFinformer对金融产品的价格预测表现出优异的性能,与其他预测方法相比,能准确捕捉长期趋势和周期性来实现高质量的预测.具体地,和传统的Transformer模型相比,在三个实验数据集上,BSFinformer的均方误差分别降低了52%,16%和19%,平均绝对误差分别降低了34%,25%和11%,为金融数据的长期时间序列预测提供了一种有效的解决方案.

浅谈数据挖掘中金融时间序列的粗糙聚类

金融时间序列分析与预测作为金融领域重要研究方向,对于揭示市场动态、指导投资决策以及维护金融稳定具有关键意义。然而,金融时间序列数据具有复杂性、高噪声等特点,使得传统聚类方法在处理这些问题上往往存在局限性。粗糙聚类作为一种基于粗糙集理论的方法,具有处理上述问题的潜力。首先介绍粗糙集理论及粗糙聚类方法的基本概念和原理。然后,重点关注粗糙聚类在金融时间序列分析与预测领域的应用,包括宏观经济预测、股票市场分析等。最后通过实际应用实例,展示粗糙聚类在金融时间序列分析与预测中的重要价值。

基于改进深度学习的金融时间序列波动率研究

线性模型和传统神经网络模型是常用的传统金融时间序列预测方法,但在非线性、非平稳的金融时间序列预测中存在一定的局限性。对此,文章提出了一种改进的深度学习模型。该模型结合了卷积神经网络和长短时记忆网络,可以有效捕捉金融时间序列中的非线性特征。通过真实数据与预测结果对比发现,文章提出的算法模型有助于捕捉金融时间序列中的非线性特征,可提高波动率预测的准确性。

时序分析下经济增长对金融衍生产品需求影响研究

为深入研究经济增长对金融衍生产品需求的影响,给金融市场参与者和投资者提供有效支持,促进金融市场的稳定发展。首先考虑了金融市场的多方面因素,建立了自回归、混合自回归和移动回归的时序分析模型。然后深入剖析了金融衍生产品的基本特征,选取了相关变量进行解释。最后说明了金融衍生产品的基本功能,了解其在经济增长下的作用。实证结果显示,非金融机构衍生产品对企业价值影响的相关系数为0.204 9,企业经济增长对衍生品需求的影响为0.000 3。通过时序分析,有助于揭示经济周期、经济波动等因素对金融衍生产品市场的影响机制,为投资者和参与者提供有效支持。

An extended sparsemax-linearmoving model with application to high-frequency financial data

There continues to be unfading interest in developing parametric max-stable processes for modelling tail dependencies and clustered extremes in time series data.However,this comes with some difficulties mainly due to the lack of models that fit data directly without transforming the data and the barriers in estimating a significant number of parameters in the existing models.In thiswork,we study the use of the sparsemaxima ofmovingmaxima(M3)process.After introducing random effects and hidden Fréchet type shocks into the process,we get an extended maxlinear model.The extended model then enables us to model cases of tail dependence or independence depending on parameter values.We present some unique properties including mirroring the dependence structure in real data,dealing with the undesirable signature patterns found in most parametricmax-stable processes,and being directly applicable to real data.ABayesian inference approach is developed for the proposed model,and it is applied to simulated and real data.

基于机器学习算法的金融市场趋势预测研究

针对当前方法对金融市场长、短期趋势的预测结果偏差较大的问题,提出了基于机器学习的金融市场趋势预测方法。依照一定时间间隔采集金融数据,产生金融数据时间序列,采用小波分析方法对金融数据时间序列进行处理,去除其中的噪声,保留近似数据。通过卷积长短期记忆神经网络对金融数据时间序列进行学习,建立金融市场预测模型。测试结果表明,所提出的方法可以有效清除噪声、平滑初始数据,在短期趋势与长期趋势预测中均具有较高预测精度。

融合双通路注意力与VT-LSTM的金融时序预测

针对现有研究对金融时序数据短期变化规律捕捉能力不足和预测精度不佳的问题,提出一种基于双通路注意力机制和改进转换门控LSTM(variant transformation-gated LSTM,VT-LSTM)的金融时序预测模型(dual-attention MDWT-CVT-LSTM)。使用多级离散小波变换(MDWT)分解股指序列得到高频和低频数据,并在融合门控单元的LSTM中加入转换门控机制,构造VT-LSTM,其能有效把控短期突变信息。在双通路注意力网络中结合VT-LSTM与一维时序卷积(Conv1D),分别提取不同频度数据的空间局部特征和时序特征,对各子序列进行预测,实现多层级多通路的预测研究。在金融股指数据集和个股数据集上对不同模型进行实验比较,结果表明提出模型预测精度优于其他方法,有良好的可行性。

基于改进随机森林算法的财务风险预警模型构建研究

财务危机是影响企业发展的重要因素,也是企业在发展中不可避免的危险因素,因此对企业财务危机进行有效预警具有重要意义。为了实现企业财务风险预测和预警,研究提出了基于K折交叉验证的随机森林算法,同时结合时间序列分析来实现更为全面的预警。结果中显示,K折交叉验证随机森林算法的指标分类精度达到了0.907,并且研究构建的财务风险预警模型在ROC曲线的性能分析中表现出较高的线下面积,即具有较高的预测性能。以上结果表明,采用K折交叉验证改进随机森林算法能够较大程度上提升预警模型的财务指标分类效果,提升预警模型的综合预警能力,为我国企业发展提供财务风险规避策略。

基于BVANet的财经新闻情感分析

股票市场的预测一直以来是金融大数据分析领域一项难题,而财经新闻中包含的内在信息对市场表现有很大影响。提出了一种基于BERT的向量自回归融合网络(BVANet),该网络通过BERT将财经新闻情感量化,后结合市场表现联合构建金融时间序列向量自回归(VAR)模型,最终实现股票的预测。结果表明,与传统算法相比,BVANet在提取新闻情绪信息和模型预测中取得了更好的效果,新闻的情绪对市场表现有预测作用。该研究可为自然语言处理在金融预测的应用提供实践参考。

金融时间序列数据可视化框架研究

针对股票软件与量化平台在数据可视化方面存在的对接困难、量化平台可视化功能不完整、独立开发可视化模块缺少参考模型等问题,建立一套金融时间序列数据可视化框架,并对框架中各模块的计算模型进行详细介绍。在多个量化平台中使用回测和模拟实盘功能对框架进行测试。结果表明,在瞬时数据量大的情况下,框架可以在两种模式下稳定运行,并且能够适应不同量化平台之间的差异,满足研究员对数据可视化的需求。

基于信号分解降噪的CNN-BiLSTM金融市场趋势预测

随着金融时间序列数据日趋复杂,如何捕捉金融数据未来多天的趋势变化成了难题。针对该问题提出了基于信号分解降噪和注意力机制的CNN-BiLSTM金融市场趋势预测模型(attention-based DWT-VMD-CBiLSTM)。首先利用离散小波变换(DWT)对原始金融股指序列进行降噪处理,然后利用变分模态分解(VMD)对降噪后的数据进一步分解为若干子序列。再结合多元基本面特征,利用基于注意力机制的CBiLSTM网络模型对各子序列进行多步预测,最后将各预测结果相加得到最终结果,实现较为长期的趋势预测。为证明所提出的模型性能,在不同金融股指数据集上与不同模型进行了实验比较。结果表明,提出的模型预测精度优于其他方法,在平均绝对误差(MAE)和平均百分比误差(MAPE)上分别达到12.28、0.39和80.27、0.71,在可决系数(R^(2))和可释方差值(EVS)上达到72%、74%和79%、69%的拟合度。

财政金融协同支农的绩效时序演进及空间收敛分析

文章采用Malmquist动态指数与空间收敛分析方法,测度分析了2010-2019年中国30个省份财政金融协同支农绩效水平的时序演进与空间收敛特征。结果表明:全国财政金融协同支农的全要素生产率指数总体呈下降趋势,2010-2019年均下降4.5%,且由东到西呈现依次递减的分布状态,存在较为明显的空间地域差异;从指数分解角度分析,技术进步指数是导致当下财政金融协同支农全要素生产率下降的主要因素;另外,全国各地区财政金融协同支农绩效水平均呈现一定的收敛特征,整体支农水平呈稳定态势发展,但地区间的收敛速度差异较为明显。因此,在政策推进过程中,应准确把握并积极响应涉农市场需求,创新支农工具和资金使用管理方式;同时贯彻落实科技兴农战略,推动农业科技成果的落地转化;另外也要强化政策辐射效应,促进要素资源的跨区域流动,实现区域协调发展。