ArticlePublisher preview available

International carbon financial market prediction using particle swarm optimization and support vector machine

To read the full-text of this research, you can request a copy directly from the authors.

Abstract and Figures

Carbon financial futures have both the characteristics of commodity futures and environmental protection and its price is affected by many factors. It is hard and complex for traditional analysis methods to get precise prediction results effectively. How to effectively predict the price trend of carbon financial futures has been focused on by both academia and traders. This study addresses the high prediction error of European allowance (EUA) futures price by constructing a novel approach by combining the support vector machine (SVM) and particle swarm optimization (PSO) algorithm. This article introduces a parameters optimization method, which provides the best parameters for SVM to improve the prediction performance of the EUA futures price. Furthermore, this research uses the realistic trading dataset containing 30,762 EUA futures closing prices to verify the effectiveness and efficiency of the PSO-SVM prediction model. The empirical results show that the prediction performance of the model, especially the radial kernel function, is significantly improved. And this approach can determine the parameters according to the characteristics of the dataset and input the parameters for training and prediction automatically. The PSO-SVM algorithm can effectively predict extreme price fluctuations and overcome the problem of high prediction error caused by parameter constraints.
This content is subject to copyright. Terms and conditions apply.
1 3
Journal of Ambient Intelligence and Humanized Computing (2022) 13:5699–5713
International carbon financial market prediction using particle swarm
optimization andsupport vector machine
JunhuaChen1· ShufanMa1· YingWu2
Received: 21 July 2020 / Accepted: 25 March 2021 / Published online: 16 April 2021
© The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2021
Carbon financial futures have both the characteristics of commodity futures and environmental protection and its price is
affected by many factors. It is hard and complex for traditional analysis methods to get precise prediction results effectively.
How to effectively predict the price trend of carbon financial futures has been focused on by both academia and traders.
This study addresses the high prediction error of European allowance (EUA) futures price by constructing a novel approach
by combining the support vector machine (SVM) and particle swarm optimization (PSO) algorithm. This article introduces
a parameters optimization method, which provides the best parameters for SVM to improve the prediction performance of
the EUA futures price. Furthermore, this research uses the realistic trading dataset containing 30,762 EUA futures closing
prices to verify the effectiveness and efficiency of the PSO-SVM prediction model. The empirical results show that the
prediction performance of the model, especially the radial kernel function, is significantly improved. And this approach can
determine the parameters according to the characteristics of the dataset and input the parameters for training and prediction
automatically. The PSO-SVM algorithm can effectively predict extreme price fluctuations and overcome the problem of high
prediction error caused by parameter constraints.
Keywords Particle swarm optimization· Support vector machine· Carbon financial futures· Parameters optimization
1 Introduction
Environmental protection has become the consensus of the
whole society. Reducing carbon emissions and promoting
the use of renewable energy are important measures to alle-
viate the currently serious environmental problems (Chen
and Wu 2020; Preethi and Tamilarasan 2020). According
to the Clean Development Mechanism (CDM) stipulated in
Article 12 of the Kyoto Protocol, carbon dioxide emissions
can be transferred between developed and developing coun-
tries to fulfill their emission reduction obligations under the
Protocol. European Union allowance (EUA) is the official
name for Europe’s emission allowances, which was defined
as the official Kyoto allowance for countries in the EUin
2008. To better measure the monetary value of EUA, the
EUA futures contracts launched by the European Energy
Exchange (EEX), Intercontinental Exchange (ICE), and the
Nasdaq have attracted the attention of many countries and
regions. EUA futures is the trading contract of EUA, which
is priced in Euros and represents 1000 units of EUA in each
minimum trading unit. The owner of each unit of EUA has
the right to emit one ton of carbon dioxide or carbon-equiv-
alent greenhouse gas (GHG). The trading members of EUA
futures are obliged to deliver or receive carbon emission
allowance to the European Union Registry corresponding
to their actual emissions in the preceding year on April 30.
With the deepening of the concept of environmental
protection, more and more traders participate in the trad-
ing of environmental futures. For the developed and devel-
oping countries, participating in EUA futures trading can
effectively avoid the economic losses caused by the sharp
fluctuations in the price of the emission allowance market.
For speculators in EUA futures trading, the high leverage
of futures can enlarge their returns. Furthermore, the price
of carbon emission allowance has a significant impact on
national economic development (Tang etal. 2020), tech-
nological progress, and the use of fossil energy. Because
* Ying Wu
1 School ofManagement Science andEngineering, Central
University ofFinance andEconomics, Beijing, China
2 Institute forSocial Development, Chinese Academy ofSocial
Sciences (CASS), Beijing, China
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
... The graphical illustration and mathematical definition of OHLC data can refer to in Section 4.1. Although there has been a substantial body of literature devoted broadly to forecasting the price of EUA futures (e.g., Tsai and Kuo, 2014;Sun et al., 2016;Zhu et al., 2018;Liu and Shen, 2020), most of these studies only focused on the forecast of the close price of EUA futures (e.g., Fan et al., 2015;Atsalakis, 2016;Zhu et al., 2017;Chen et al., 2021). In contrast, there is a shortage of studies concerned with the structural forecasting of EUA futures price by treating the open-high-low-close prices of EUA futures as an integral data structure and jointly considering all four-dimensional prices. ...
... The results showed that the forecast accuracy of the EWT-GRU model was better than ARIMA, BPNN, GRU, and EWT-BPNN. Chen et al. (2021) addressed the high forecast error of EUA futures price by constructing a novel approach by combining SVM and particle swarm optimization (PSO) algorithms. Results indicated that the PSO-SVM algorithm could effectively forecast extreme price fluctuations and overcome high forecast errors caused by parameter constraints. ...
This paper develops an open-high-low-close (OHLC) data forecasting framework to forecast EUA futures price based on EU ETS data and extended exogenous variables from 2013 to 2020. The challenge of forecasting such an OHLC structure lies in handling its three intrinsic constraints, i.e., the positive constraint, interval constraint, and boundary constraint. This paper proposes a novel unconstrained transformation method for OHLC data and combines it with various forecasting models. Out-of-sample modelings identify the extraordinary performance of the convolutional neural network (CNN) in terms of MAPE (1.371%), MAE (0.274), RMSE (0.370), and AR (0.621), better than that of multiple linear regression (MLR), vector auto-regression (VAR) and vector error correction model (VECM), support vector regression (SVR), and multi-layer perceptron (MLP). The proposed transformation-based forecasting framework demonstrates the considerable potential for OHLC data forecasting in the energy finance field, e.g., crude and natural gas. Practicable and concrete suggestions are provided to ensure the profitability of trading EUA futures.
... For example, Tang et al. (Tang et al., 2020) mention that the integration of artificial intelligence (AI) models and statistical models can improve the accuracy of the interpretability of carbon market research. Chai et al. and Chen et al.(Chai et al., 2020;Chen et al., 2021) use support vector machine (SVM) and particle swarm optimization (PSO) techniques, respectively, to forecast carbon prices, helping investors and risk managers in their decision-making. Therefore, future research should balance adequate models for carbon market-related research. ...
Full-text available
Green finance is an emerging topic which is broadly discussed in context of adapting and mitigating environmental deterioration due to climate change. As an effective incentive mechanism, it provides strong support for carbon emission reduction. However, a limited review articles investigate the specific combination of green finance and carbon emission reduction. Here, we apply a bibliometric analysis to review research on green finance and carbon emission reduction based on the literature from 2010 to 2021 in the Web of Science core database. The results indicate that countries with the most publications were those with high economic development, salient environmental problems, and a strong demand for ecological protection. Top publishing journals include Climate Policy, Journal of Cleaner Production, and Energy Policy. The author collaboration is fragmented, mostly less than three researchers. Based on analyses of keyword frequency and centrality, deforestation, carbon markets, and financial development were the most significant research topics. The research hotspots included clean development mechanism, adaptation, carbon market, and sequestration. Finally, the DPSIR framework is applied to explore driving forces, state, pressure, impact and response of current research. We hope our work provides a systematic review of green finance for carbon emission reduction to boost the research in this field.
... By contrast with econometric models, machine learning models are effective in mining and retaining the valuable information in the data and in dealing with nonlinear data effectively [7]. Artificial neural network (ANN) [8,9], support vector machine (SVM) [10,11], long short-term memory (LSTM) [12], and ensemble learning models are some of the common models. Ensemble learning models can effectively combine the results of multiple base learners to achieve secondary learning of the problem with high generalizability. ...
Full-text available
An integrated hog futures price forecasting model based on whale optimization algorithm (WOA), LightGBM, and Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) is proposed to overcome the limitations of a single machine learning model with low prediction accuracy and insufficient model stability. The simulation process begins with a grey correlation analysis of the hog futures price index system in order to identify influencing factors; after that, the WOA-LightGBM model is developed, and the WOA algorithm is used to optimize the LightGBM model parameters; and, finally, the residual sequence is decomposed and corrected by using the CEEMDAN method to build a combined WOA-LightGBM-CEEMDAN model. Furthermore, it is used for comparison experiments to check the validity of the model by using data from CSI 300 stock index futures. Based on all experimental results, the proposed combined model shows the highest prediction accuracy, surpassing the comparative model. The model proposed in this study is accurate enough to meet the forecasting accuracy requirements and provides an effective method for forecasting future prices.
Full-text available
Carbon trading price (CTP) prediction accuracy is critical for both market participants and policymakers. As things stand, most previous studies have only focused on one or a few carbon trading markets, implying that the models’ universality is insufficient to be validated. By employing a case study of all carbon trading markets in China, this study proposes a hybrid point and interval CTP forecasting model. First, the Pearson correlation method is used to identify the key influencing factors of CTP. The original CTP data is then decomposed into multiple series using complete ensemble empirical mode decomposition with adaptive noise. Following that, the sample entropy method is used to reconstruct the series to reduce computational time and avoid overdecomposition. Following that, a long short-term memory method optimized by the Adam algorithm is established to achieve the point forecasting of CTP. Finally, the kernel density estimation method is used to predict CTP intervals. On the one hand, the results demonstrate the proposed model’s validity and superiority. The interval prediction model, on the other hand, reflects the uncertainty of market participants’ behavior, which is more practical in the operation of carbon trading markets. Graphical abstract
Faults are most likely to occur during the coding phase of software development. If, before the testing process, we can predict parts of code that are more prone to faults, then a large amount of time, software cost could be saved, and the software’s overall quality could be improved. Various researchers have previously attempted to predict software faults using numerous machine learning techniques in order to identify whether software modules are fault-prone or not. Ranking the software modules based on their fault content has rarely been explored before. Additionally, Bayesian methods have not been explored before for this task. We aim to investigate both linear and non-linear Bayesian regression methods for software fault prediction in this work. We develop and evaluate fault prediction models for two scenarios: intra-release prediction and cross-release prediction. The experimental investigation is conducted on 46 different software project versions. We use mean absolute error, and root means square error, and fault percentage average as performance measures. The results showed that Bayesian NLR outperformed linear regression and other used machine learning approaches or produced at least comparable performance. Bayesian linear regression method performed moderately.
Full-text available
Digital currency price prediction is vital to both sellers and purchasers. Over these years, decomposition and integration models have been applied more and more to realize the goal of precise prediction, however, many of them tend to neglect the reconstruction of features or the residual series. Altogether, one of the biggest drawbacks of the decomposition and integration framework is the method applied requires manual parameter setting whether it is for decomposition or integration. Still, for the results, they are merely satisfied with the point prediction which brings high uncertainty. In this paper, an optimized feature reconstruction decomposition and two-step nonlinear integration method is proposed which gives consideration to feature reconstruction, nonlinear integration, optimization and interval prediction. The original data series is decomposed through improved variational mode decomposition based approximate entropy feature reconstruction system. Then, improved particle swarm optimization-gated recurrent unit (iPSO-GRU) is utilized in the first and second nonlinear integration part separately. Meanwhile, the residual series is given attention, if it is not a white noise series, the residual will be the input of iPSO-GRU whose result will be added back to the second integration result to form the point prediction result. Based on the point prediction result, interval prediction estimate will be generated as well via maximum likelihood function. This study chooses three kinds of digital currency as cases and the results show that the MAPE values of point prediction are all below 3.5%, and CP values of interval prediction are all 1 with suitable MWP. In addition, compared with other benchmark models, the proposed model shows better performance.
Full-text available
An efficient air pollution monitoring (APM) scheme is proposed to establish an optimal less polluted path using WSN (wireless sensor network), in which the sensor node (SN) senses the temperature and CO gas (carbon monoxide) concentration’s existent in the air. Initially, the sensed information from the SNs is preprocessed. During preprocessing, the value that is missed in the sensed information is imputed. Next, Hadoop's distributed file system (HDFS) MapReduce (MR) is implemented on the preprocessed data and subsequently, the resulting data is saved in the cloud server. The resulting data is analyzed using Improved-Adaptive Neuro-Fuzzy Inference System (I-ANFIS) Algorithm for checking air pollutions severities and its location is then presented in the Google Map. After that, the multi-path routing is established through the less polluted area. Lastly, the optimal path is chosen with the assistance of KHOA (Krill Herd Optimization Algorithm). The outcomes are evaluated by contrasting the proposed and prevailing techniques.
Full-text available
The target of flood hazard assessment is to establish appropriate casualty evaluation models for managing flood and alleviating loss affected by flood. Existing challenges on evaluating flood hazard level focus on how to achieve a stable and high-resolution result according to the incompatibilities between multiple attribute indicators. Base on the chaotic optimization theory, this study presents a chaotic particle swarm optimization (CPSO) algorithm to deal with a fuzzy clustering iterative model (CPSO-FCI) for evaluating flood hazards. By using a celebrated logistic chaotic map and a penalty function, the target function can be tackled more perfectly. The effectiveness of the novel hybrid method is evaluated by three representative test functions and two sets of practical flood disaster samples in China. Simulation results and comparisons show that the presented CPSO based on the logistic map is competitive and stable in performance with standard particle swarm optimization (PSO) and other advanced PSO-type approaches introduced in the literature. High resolution continuous flood rating results are obtained for fine-tune flood management.
Full-text available
Together with water and land, air is a fundamental necessity of life. Nevertheless, the ambient air quality is deteriorating around the world because of rapid urbanization and industrialization. The problem of air pollution has become a prominent issue for the public and academia. In fact, the public is more interested in being informed about the possibility of occurrence of air pollution episodes than the accurate forecasting of a specific pollutant. Therefore, this study proposes a process based upon Discrete-Time Markov Chains (DTMC), to predict the air quality index (AQI) and identify the prime air pollutants in a specific area. This study utilizes online air quality monitoring data retrieved from the Taiwan Environment Protection Administration, to demonstrate the application of the process. The findings of the study revealed that there are three prime air pollutants, namely ozone (O3), nitrogen dioxide (NO2), and fine particulate matter (PM10), which frequently contaminate the ambient air in Taipei city. Furthermore, this study used data for three time periods to verify the proposed process and found that the performance of the process in predicting the AQI values for seven days is better than the prediction for 30 days and 62 days.
Full-text available
The rainfall prediction is important for metrological department as it closely associated with our environment and human life. An accuracy of rainfall prediction has great important for countries like India whose economy is dependent on agriculture. Because of dynamic nature of atmosphere, statistical techniques fail to predict rainfall information. The process of support vector machine (SVM) is to find an optimal boundary also known as hyper plane in which separates the samples (examples in a dataset) of different classes by a maximum margin. The proposed model uses the dynamic integrated model for exploring and learning large amount of data set. Balanced communication-avoiding support vector machine (CA-SVM) prediction model is proposed to achieve better performance and accuracy with limited number of iteration without any error. The rain fall dataset is used for performance evaluation. The proposed model starts with independent sample to the integrated samples without any collision in prediction. The proposed algorithm achieves 89% of accuracy when compared to the existing algorithms. The simulations demonstrate that prediction models indicate that the performance of the proposed algorithm Balanced CA-SVM has much better accuracy than the local learning model based on a set of experimental data if other things are equal. On the other hand, simulation results demonstrate the effectiveness and advantages of the Balanced CA-SVM model used in machine learning and further promises the scope for improvement as more and more relevant attributes can be used in predicting the dependent variables.
This study extracts the hidden topic model and emotional information from news articles. A novel fuzzy twin support vector machine is also developed to merge the large volume of information from on-line news, using this to predict stock price trends. Fuzzy set theory is very useful for this approach because the texts use fuzzy terminology, such as high/low and big/small, and the boundary between rising and falling categories is poorly defined. By using fuzzy partial ordering relation, the decision function of our approach is generalized such that the values assigned to the stocks fall within a specified range and indicate the membership grade of these stocks in the positive class (rising trends). The proposed approach is useful for dealing with outliers. Outliers only slightly increase the spread of the fuzzy decision boundary and the estimation of the mode of fuzzy hyperplane remains unchanged. Therefore, our model is more robust compared with other methods when the data contain some outliers. Besides, the membership grade estimated by the fuzzy decision boundary gives a better insight into the degree of confidence in the predicted outputs. The explicable characteristic for the degree of confidence is a vital aspect for decision-making applications.
Icing of wind turbine blades is a phenomenon that commonly occurs in autumn and winter and critically affects the safety and efficiency of wind turbine operation. Therefore, it is of considerable significance to predict whether wind turbine blades are frozen. The data used in this work are obtained from a supervisory control and data acquisition system. The particle swarm optimization algorithm is used to optimize the kernel function of the support vector machine to establish a model to predict whether a fan blade is frozen. Specifically, first, the data are preprocessed to eliminate apparent ice free data, and the data sets are further balanced using undersampling and oversampling techniques. Second, the appropriate eigenvalues are selected according to the icing mechanism. The optimal parameters of the support vector machine are obtained using particle swarm optimization algorithm. Finally, the characteristic value and parameters are substituted into the support vector machine to evaluate the fault mechanism of blade icing.
Wind power output is highly volatile and intermittent owing to the characteristics of wind energy. Large-scale wind power integration affects the stability of entire power system. Accurate wind power prediction facilitates to utilize wind energy, improve power supply quality and maintain stable operation of power grid. A hybrid prediction model combining wavelet decomposition, support vector machine and improved atom search algorithm is proposed to predict wind power output. Dynamic sinusoidal wave adaptive weight is introduced to improve the position update equation, and crossover and mutation operations are added to the end of each iteration to improve the search ability of atomic search algorithm. In the prediction process, the original wind power data is processed by wavelet decomposition and the non-stationary signal is decomposed into several detail sequences of different frequencies and approximate sequence to extract important wind power features to reduce the prediction error caused by data fluctuation. Prediction results of actual wind farm data prove the proposed model has prominent advantages in predicting performance. Three evaluation indexes of mean absolute error, mean absolute percent error and root mean square error decreased by at least 1.14%, 2.60% and 1.52%. The proposed method is conductive to reduce wind curtailment and improve the economic benefit of wind farm.
The European Union Emissions Trading Scheme (EU ETS) was created in 2005 to price every ton of carbon dioxide emissions. Within this framework, EU carbon dioxide emission allowances can affect electric power industry stock performance. This article uses a multifactor market model and a panel data econometric technique to investigate the long-run impact of EU carbon dioxide emission allowances on the European power sector. We also use panel cointegration to check whether there is a long-run relationship, and fully modified ordinary least square and dynamic ordinary least square to estimate any such relationship. The panel data include a daily sample for the ongoing EU ETS Phase III (from 1 January 2013 until 22 April 2017) and data from six European Union members (Austria, France, Germany, Italy, Netherlands, and Spain). The estimated coefficients suggest that EU allowance prices have a statistically significant and positive long-run effect on the European power sector stock market in EU ETS Phase III. This potentially supports EU efforts to toughen carbon reduction regime targets in order to remove the surplus from the system.
Bitcoin as an asset class has received phenomenal investor attention and is considered to have similar characteristics like gold. This study aims to analyze the price behavior of bitcoin and apply machine learning algorithm for its prediction. Understanding the nature of Bitcoin price series is a multi-scale problem, and it can be best examined by analyzing its compositional characteristics. This study uses complete empirical ensemble mode decomposition (CEEMD) to analyze the nature of Bitcoin price series. Daily Bitcoin prices from 2012 to 2018 are used to perform CEEMD to identify the short term, medium term, and long-term trend in the Bitcoin price series. The study uses support vector machine (SVM) learning algorithm to find whether it can predict Bitcoin prices and finds that SVM predicts five steps ahead Bitcoin prices for the short term, medium term, long term, and overall Bitcoin price level.
Long-span spatial steel structures are affected by unfavourable factors during its service life, which can cause damage accumulation, and consequently hazards. The monitoring data of structural response of spatial steel is the basis for understanding structural behaviour and assessing structural safety, but limited number and location of sensors lead to incomplete response acquisition. Because structural members are connected to each other by joints, their structural responses are correlated to a certain degree. Based on this fact, in this paper, a structural response estimation method based on the mapping relationship and a correlation analysis is proposed. The mapping relationship between the measured response and the estimated response is determined by support vector machine with particle swarm optimisation under any load. From this, the method to obtain the optimal mapping relationship is established to realize the response estimation under different loads. In other words, through the correlation analysis of structural response under different loads, the measured point and its corresponding estimated points are determined according to the law of variation of the correlation, and parameters of the optimal mapping model are obtained. Based on the stress and displacement responses of the steel roof above the Shenzhen bay stadium under different wind loads, the effectiveness and feasibility of the proposed method are verified.