A Novel Fuzzy Linear Regression Sliding Window GARCH Model for Time-Series Forecasting

Generalized autoregressive conditional heteroskedasticity (GARCH) is one of the most popular models for time-series forecasting. The GARCH model uses a maximum likelihood method for parameter estimation. For the likelihood method to work, there should be a known and specific distribution. However, due to uncertainties in time-series data, a specific distribution is indeterminable. The GARCH model is also unable to capture the influence of each variance in the observation because the calculation of the long-run average variance only considers the series in its entirety, hence the information on different effects of the variances in each observation is disregarded. Therefore, in this study, a novel forecasting model dubbed a fuzzy linear regression sliding window GARCH (FLR-FSWGARCH) model was proposed; a fuzzy linear regression was combined in GARCH to estimate parameters and a fuzzy sliding window variance was developed to estimate the weight of a forecast. The proposed model promotes consistency and symmetry in the parameter estimation and forecasting, which in turn increases the accuracy of forecasts. Two datasets were used for evaluation purposes and the result of the proposed model produced forecasts that were almost similar to the actual data and outperformed existing models. The proposed model was significantly fitted and reliable for time-series forecasting.

Download Full-text

A Novel Fuzzy Time Series Forecasting Model Based on Multiple Linear Regression and Time Series Clustering

Mathematical Problems in Engineering ◽

10.1155/2020/9546792 ◽

2020 ◽

Vol 2020 ◽

pp. 1-17

Author(s):

Yanpeng Zhang ◽

Hua Qu ◽

Weipeng Wang ◽

Jihong Zhao

Keyword(s):

Time Series ◽

Linear Regression ◽

Multiple Linear Regression ◽

Time Series Forecasting ◽

High Order ◽

Fuzzy Time Series ◽

Forecasting Model ◽

Time Series Clustering ◽

Proposed Model ◽

High Order Time

Time series forecasting models based on a linear relationship model show great performance. However, these models cannot handle the the data that are incomplete, imprecise, and ambiguous as the interval-based fuzzy time series models since the process of fuzzification is abandoned. This article proposes a novel fuzzy time series forecasting model based on multiple linear regression and time series clustering for forecasting market prices. The proposed model employs a preprocessing to transform the set of fuzzy high-order time series into a set of high-order time series, with synthetic minority oversampling technique. After that, a high-order time series clustering algorithm based on the multiple linear regression model is proposed to cluster dataset of fuzzy time series and to build the linear regression model for each cluster. Then, we make forecasting by calculating the weighted sum of linear regression models’ results. Also, a learning algorithm is proposed to train the whole model, which applies artificial neural network to learn the weights of linear models. The interval-based fuzzification ensures the capability to deal with the uncertainties, and linear model and artificial neural network enable the proposed model to learn both of linear and nonlinear characteristics. The experiment results show that the proposed model improves the average forecasting accuracy rate and is more suitable for dealing with these uncertainties.

Download Full-text

A Neutrosophic Forecasting Model for Time Series Based on First-Order State and Information Entropy of High-Order Fluctuation

Entropy ◽

10.3390/e21050455 ◽

2019 ◽

Vol 21 (5) ◽

pp. 455 ◽

Cited By ~ 5

Author(s):

Hongjun Guan ◽

Zongli Dai ◽

Shuang Guan ◽

Aiwu Zhao

Keyword(s):

Time Series ◽

Information Entropy ◽

Stock Exchange ◽

Time Series Forecasting ◽

High Order ◽

Stock Index ◽

Neutrosophic Set ◽

Forecasting Model ◽

Forecasting Models ◽

Proposed Model

In time series forecasting, information presentation directly affects prediction efficiency. Most existing time series forecasting models follow logical rules according to the relationships between neighboring states, without considering the inconsistency of fluctuations for a related period. In this paper, we propose a new perspective to study the problem of prediction, in which inconsistency is quantified and regarded as a key characteristic of prediction rules. First, a time series is converted to a fluctuation time series by comparing each of the current data with corresponding previous data. Then, the upward trend of each of fluctuation data is mapped to the truth-membership of a neutrosophic set, while a falsity-membership is used for the downward trend. Information entropy of high-order fluctuation time series is introduced to describe the inconsistency of historical fluctuations and is mapped to the indeterminacy-membership of the neutrosophic set. Finally, an existing similarity measurement method for the neutrosophic set is introduced to find similar states during the forecasting stage. Then, a weighted arithmetic averaging (WAA) aggregation operator is introduced to obtain the forecasting result according to the corresponding similarity. Compared to existing forecasting models, the neutrosophic forecasting model based on information entropy (NFM-IE) can represent both fluctuation trend and fluctuation consistency information. In order to test its performance, we used the proposed model to forecast some realistic time series, such as the Taiwan Stock Exchange Capitalization Weighted Stock Index (TAIEX), the Shanghai Stock Exchange Composite Index (SHSECI), and the Hang Seng Index (HSI). The experimental results show that the proposed model can stably predict for different datasets. Simultaneously, comparing the prediction error to other approaches proves that the model has outstanding prediction accuracy and universality.

Download Full-text

Forecasting stock market price by using fuzzified Choquet integral based fuzzy measures with genetic algorithm for parameter optimization

RAIRO - Operations Research ◽

10.1051/ro/2019117 ◽

2020 ◽

Vol 54 (2) ◽

pp. 597-614

Author(s):

Shanoli Samui Pal ◽

Samarjit Kar

Keyword(s):

Genetic Algorithm ◽

Time Series ◽

Linear Regression ◽

Regression Model ◽

Choquet Integral ◽

Time Series Forecasting ◽

Fuzzy Measure ◽

Model Parameters ◽

Forecasting Models ◽

Non Linear

In this paper, fuzzified Choquet integral and fuzzy-valued integrand with respect to separate measures like fuzzy measure, signed fuzzy measure and intuitionistic fuzzy measure are used to develop regression model for forecasting. Fuzzified Choquet integral is used to build a regression model for forecasting time series with multiple attributes as predictor attributes. Linear regression based forecasting models are suffering from low accuracy and unable to approximate the non-linearity in time series. Whereas Choquet integral can be used as a general non-linear regression model with respect to non classical measures. In the Choquet integral based regression model parameters are optimized by using a real coded genetic algorithm (GA). In these forecasting models, fuzzified integrands denote the participation of an individual attribute or a group of attributes to predict the current situation. Here, more generalized Choquet integral, i.e., fuzzified Choquet integral is used in case of non-linear time series forecasting models. Three different real stock exchange data are used to predict the time series forecasting model. It is observed that the accuracy of prediction models highly depends on the non-linearity of the time series.

Download Full-text

An Efficient Time Series Forecasting Method Exploiting Fuzziness and Turbulences in Data

Intelligent Systems ◽

10.4018/978-1-5225-5643-5.ch078 ◽

2018 ◽

pp. 1773-1791 ◽

Cited By ~ 1

Author(s):

Prateek Pandey ◽

Shishir Kumar ◽

Sandeep Shrivastava

Keyword(s):

Time Series ◽

Performance Comparison ◽

Time Series Forecasting ◽

Fuzzy Time Series ◽

Data Set ◽

Common Trend ◽

Proposed Model ◽

Forecasting Method ◽

High Degree ◽

Degree Of Instability

In recent years, there has been a growing interest in Time Series forecasting. A number of time series forecasting methods have been proposed by various researchers. However, a common trend found in these methods is that they all underperform on a data set that exhibit uneven ups and downs (turbulences). In this paper, a new method based on fuzzy time-series (henceforth FTS) to forecast on the fundament of turbulences in the data set is proposed. The results show that the turbulence based fuzzy time series forecasting is effective, especially, when the available data indicate a high degree of instability. A few benchmark FTS methods are identified from the literature, their limitations and gaps are discussed and it is observed that the proposed method successfully overcome their deficiencies to produce better results. In order to validate the proposed model, a performance comparison with various conventional time series models is also presented.

Download Full-text

Enhanced Forecasting Accuracy of Fuzzy Time Series Model Based on Combined Fuzzy C-Mean Clustering with Particle Swam Optimization

International Journal of Computational Intelligence and Applications ◽

10.1142/s1469026820500170 ◽

2020 ◽

Vol 19 (02) ◽

pp. 2050017

Author(s):

Nghiem Van Tinh

Keyword(s):

Time Series ◽

Historical Data ◽

Pso Algorithm ◽

Time Series Forecasting ◽

Fuzzy Time Series ◽

Forecasting Accuracy ◽

Forecasting Models ◽

Proposed Model ◽

Fuzzy C Means Clustering ◽

Universe Of Discourse

Over the past 25 years, numerous fuzzy time series forecasting models have been proposed to deal the complex and uncertain problems. The main factors that affect the forecasting results of these models are partition universe of discourse, creation of fuzzy relationship groups and defuzzification of forecasting output values. So, this study presents a hybrid fuzzy time series forecasting model combined particle swarm optimization (PSO) and fuzzy C-means clustering (FCM) for solving issues above. The FCM clustering is used to divide the historical data into initial intervals with unequal size. After generating interval, the historical data is fuzzified into fuzzy sets with the aim to serve for establishing fuzzy relationship groups according to chronological order. Then the information obtained from the fuzzy relationship groups can be used to calculate forecasted value based on a new defuzzification technique. In addition, in order to enhance forecasting accuracy, the PSO algorithm is used for finding optimum interval lengths in the universe of discourse. The proposed model is applied to forecast three well-known numerical datasets (enrolments data of the University of Alabama, the Taiwan futures exchange —TAIFEX data and yearly deaths in car road accidents in Belgium). These datasets are also examined by using some other forecasting models available in the literature. The forecasting results obtained from the proposed model are compared to those produced by the other models. It is observed that the proposed model achieves higher forecasting accuracy than its counterparts for both first—order and high—order fuzzy logical relationship.

Download Full-text

A new model for learning-based forecasting procedure by combining k-means clustering and time series forecasting algorithms

PeerJ Computer Science ◽

10.7717/peerj-cs.534 ◽

2021 ◽

Vol 7 ◽

pp. e534

Author(s):

Kristoko Dwi Hartomo ◽

Yessica Nataliani

Keyword(s):

Time Series ◽

Clustering Algorithm ◽

Time Series Data ◽

Mean Squared Error ◽

Time Series Forecasting ◽

Series Data ◽

Improvement Rate ◽

New Model ◽

Average Improvement ◽

Proposed Model

This paper aims to propose a new model for time series forecasting that combines forecasting with clustering algorithm. It introduces a new scheme to improve the forecasting results by grouping the time series data using k-means clustering algorithm. It utilizes the clustering result to get the forecasting data. There are usually some user-defined parameters affecting the forecasting results, therefore, a learning-based procedure is proposed to estimate the parameters that will be used for forecasting. This parameter value is computed in the algorithm simultaneously. The result of the experiment compared to other forecasting algorithms demonstrates good results for the proposed model. It has the smallest mean squared error of 13,007.91 and the average improvement rate of 19.83%.

Download Full-text

ỨNG DỤNG ĐẠI SỐ GIA TỬ TRONG DỰ BÁO CHUỖI THỜI GIAN MỜ

Vietnam Journal of Science and Technology ◽

10.15625/0866-708x/54/2/5242 ◽

2016 ◽

Vol 54 (2) ◽

pp. 161

Author(s):

Nguyễn Cát Hồ ◽

Nguyễn Công Điều ◽

Vũ Như Lân

Keyword(s):

Time Series ◽

Time Series Forecasting ◽

Fuzzy Time Series ◽

Fuzzy Approach ◽

University Of Alabama ◽

Simple Arithmetic ◽

Proposed Model ◽

Improving Accuracy ◽

The University ◽

And Control

Fuzzy time series given by Song & Chissom (1993) in magazine "Fuzzy Sets and Systems" has been widely studied for forecasting purposes. However, the accuracy of forecasts based on the concept of fuzzy approach of Song & Chissom is not high because of such depends on many factors. Chen (1996) proposed an efficient fuzzy time series model which consists of simple arithmetic calculations only. After that, this has been widely studied for improving accuracy of forecasting in many applications to get better results. The hedge algebras developed by Nguyen and Wechler (1990) was completely different from the fuzzy approach. Here the hedge algebras was used to model linguistic domains and variables and their semantic structure is obtained. Instead of performing fuzzification and defuzzification, more simple methods are adopted, termed as semantization and desemantization, respectively. The hedge algebras based fuzzy system is a new topic, which was first applied to fuzzy control 2008 [16]. Hedge algebras applications for some specific problems in the field of information technology and control has a number of important results and confirm advantages of this approach in comparing with fuzzy approach. In continuilty of hedge algebras applications, this paper is mainly focused on the field of fuzzy time series forecasting under hedge algebras approach. In this paper, we present a new approach using hedge algebras to provide a computational model, which is completely different from the fuzzy approach for fuzzy time series forecasting. The experimental results of forecasting enrollments of students of the University of Alabama show that the model of fuzzy time series based on hedge algebras is better than many existing models. We can see that the proposed model gains higher forecasting accuracy than the original model presented by Song and Chissom (1993b), Chen (1996, 2002), or Lee (2009), Qiu (2011), Egrioglu (2012), Ozdemir ( 2012) and Uslu (2013).

Download Full-text

A hybrid modelling method for time series forecasting based on a linear regression model and deep learning

Applied Intelligence ◽

10.1007/s10489-019-01426-3 ◽

2019 ◽

Vol 49 (8) ◽

pp. 3002-3015 ◽

Cited By ~ 6

Author(s):

Wenquan Xu ◽

Hui Peng ◽

Xiaoyong Zeng ◽

Feng Zhou ◽

Xiaoying Tian ◽

...

Keyword(s):

Time Series ◽

Deep Learning ◽

Linear Regression ◽

Regression Model ◽

Linear Regression Model ◽

Time Series Forecasting ◽

Hybrid Modelling ◽

Modelling Method

Download Full-text

Aplikasi GARCH dalam Mengatasi Volatilitas Pada Data Keuangan

Jurnal Matematika ◽

10.24843/jmat.2017.v07.i02.p87 ◽

2017 ◽

Vol 7 (2) ◽

pp. 107

Author(s):

, Hartati ◽

Imelda Saluza

Keyword(s):

Time Series ◽

Garch Model ◽

Practical Experience ◽

Financial Data ◽

Demand And Supply ◽

Volatility Clustering ◽

Wide Range ◽

One Year ◽

The Garch Model

The financial market is a place or means convergence between demand and supply of a wide range of financial instruments Long-term (over one year). Activities that occur in the financial markets in the long term will form a series of data is often called a time series that contains a set of information from time to time. Practical experience shows that many time series exhibit their periods with great volatility. The greater the volatility, the greater the chance to experience a gain or loss. Important properties are often owned by the data time series in finance, especially to return data that the probability distribution of returns are fat tails (tail fat) and volatility clustering or often referred to as a case heteroskedastisitas. Not all models are able to capture the nature of heteroscedasticity, one of the models that are able to do is Generalized Autoregressive Heteroskedasticity Condition (GARCH). So the purpose of this study was to determine the GARCH model in dealing with the volatility that occurred in the financial data. The results showed that the GARCH model is best suited to see volatility in the financial data.

Download Full-text

A GOAL PROGRAMMING APPROACH TO FUZZY LINEAR REGRESSION WITH NON-FUZZY INPUT AND FUZZY OUTPUT DATA

Asia Pacific Journal of Operational Research ◽

10.1142/s0217595909002420 ◽

2009 ◽

Vol 26 (05) ◽

pp. 587-604 ◽

Cited By ~ 12

Author(s):

H. HASSANPOUR ◽

H. R. MALEKI ◽

M. A. YAGHOOBI

Keyword(s):

Linear Regression ◽

Goal Programming ◽

Programming Approach ◽

Simulation Studies ◽

Fuzzy Linear Regression ◽

Output Data ◽

The Past ◽

Proposed Model ◽

Fuzzy Input ◽

Fuzzy Linear Regression Model

Many researches have been carried out in fuzzy linear regression since the past three decades. In this paper, a fuzzy linear regression model based on goal programming is proposed. The proposed model takes into account the centers of fuzzy data as an important feature as well as their spreads. Furthermore, the model can deal with both symmetric and non-symmetric data. To show the efficiency of proposed model, it is compared with some earlier methods based on simulation studies and numerical examples. Moreover, the sensitivity of the model to outliers is discussed.

Download Full-text