Fuzzy Time Series Forecasting Research Articles

A study and development of high‐order fuzzy time series forecasting methods for air quality index forecasting

AbstractThe endless adverse effects of air pollution incidents have raised significant public concerns in the past few decades. The measure of air pollution, that is, the air quality index (AQI), is highly volatile and associated with different kinds of uncertainties. Following this, the study and development of accurate fuzzy time series forecasting (TSF) methods for predicting the AQI have a significant role in air pollution control and management. Motivated by this, in this paper, a systematic study is made to evaluate the true potential of fuzzy TSF methods employing traditional fuzzy set (TFS), intuitionistic fuzzy set (IFS), hesitant fuzzy set (HFS), and neutrosophic fuzzy set (NFS) in forecasting the AQI. Two novel high‐order fuzzy TSF methods, TFS‐multilayer perceptron (MLP) and HFS‐MLP, are proposed employing TFS and HFS in which ratio trend variation of AQI data is used instead of original AQI, MLP is used to model the fuzzy logical relationships (FLRs), and none/mean of aggregated membership values are used while modeling the FLRs using MLP. The results from the proposed fuzzy TSF methods are compared with recently proposed fuzzy TSF methods employing TFS, IFS, and NFS and six popular machine learning models, including MLP, support vector regression (SVR), Bagging Regressors, XGBoost, long‐short term memory (LSTM), and convolutional neural network (CNN). The “Wilcoxon Signed‐Rank test” and “Friedman and Nemenyi hypothesis test” are applied to the results obtained by employing different ratios in the train‐validation‐test to draw decisive conclusions reliably. The simulation results show the statistical dominance of the proposed TFS‐MLP method over all other crisp and fuzzy TSF methods employed in this paper.

Intuitionistic fuzzy time series forecasting method based on dendrite neuron model and exponential smoothing

Methods based on artificial neural networks for intuitionistic fuzzy time series forecasting can produce successful forecasting results. In the literature, exponential smoothing methods are hybridised with artificial neural networks due to their simple and efficient structures to improve the forecasting performance. The contribution of this paper is to propose a new forecasting approach combining exponential smoothing methods and intuitionistic fuzzy time series. In this study, a forecasting algorithm based on the dendrite neuron model and simple exponential smoothing methods is proposed for modelling intuitionistic fuzzy time series. In the fuzzification stage of the proposed method, the intuitionistic fuzzy c-means method is used. The proposed method is a modular method using two separate dendrite neuron model neural networks and the grey wolf optimisation algorithm is used to estimate all parameters of the method. The performance of the proposed method is applied on four different random time series obtained for Index of Coin Market Cap and the performance of the method is compared with some other fuzzy forecasting methods. As a result of the analyses, it is concluded that the proposed modular method has better forecasting results than other methods.

Singh's Fuzzy Time Series Forecasting Modification Based on Interval Ratio

Background: One forecasting method that is often used is time series forecasting. The development of applied mathematics has encouraged new mathematical findings that led to the birth of new branches of mathematics, one of which is fuzzy. Purpose: The objectives of the study, namely forecasting, fuzzy set, time series, fuzzy time series, fuzzy time series Singh, interval ratio and measurement of accuracy level. Method: This research method applies Chen's fuzzy time series in the section of determining the universe of talk you to the fuzzification of historical data and in the part of forecasting results obtained through a heuristic approach by building three forecasting rules, namely Rule 2.1, Rule 2.2, and Rule 2.3 to obtain better results and affect very small AFER values. As well as making modifications to the interval partition section using interval ratios to be able to reflect data variations. Results: Based on the calculation of AFER values for order 2, order 3, and order 4 respectively obtained at 1.06389%, 0.689368%, and 0.711947%. Therefore, it can be said, Singh's fuzzy time series forecasting method based on the ratio of 3rd-order intervals is better than that of 2nd-order and 4th-order. Conclusion: Based on the results of research and discussion that has been carried out, it can be concluded that Singh's fuzzy time series forecasting method has the same algorithm as fuzzy time series forecasting. Singh's fuzzy time series forecasting method based on interval ratios applies fuzzy time series and Singh forecasting. Singh's fuzzy time series forecasting modification accuracy rate based on interval ratios produces excellent forecasting values according to evaluator average forecasting error rate (AFER).

Differential Convolutional Fuzzy Time Series Forecasting

Differential Convolutional Fuzzy Time Series Forecasting

Strong (α, k)-cut and computational-based segmentation based novel hesitant fuzzy time series forecasting model

Hesitant fuzzy set (HFS) is an ideal tool to deal with hesitancy situation arises due to collection of membership values for each time series datum. Many researchers developed HFS based fuzzy time series forecasting (FTSF) models in past years. In this article, a novel method is proposed to deal with critical factors affecting the HFS based FTSF and name as strong (α,k)-cut and computational-based segmentation based hesitant fuzzy time series forecasting (SCS-FTSF). Computational-based segmentation (CBS) approach is developed to determination of number of intervals and generating intervals. Proposed method uses Gaussian and triangular membership functions to construct HFS and uses aggregation operator aggregating the membership values to construct aggregate HFS. A novel fuzzification procedure is proposed by taking all aggregate HFSs with non-zero aggregated membership value relative to data points. Strong(α,k)-cut is employed to selection of suitable aggregate HFLRs. Further, a defuzzification approach is also proposed to obtaining numerical value. In order to assess the performance of proposed SCS -FTSF method two time series datasets are used. Results of error measures and validation tests confirms that superiority of the proposed SCS-FTSF method.

Multi-variable time series forecasting model based on high-order hesitant probabilistic linguistic fuzzy logical relationship

The stock market is uncertain, but its fluctuations have inherent laws. A suitable method to extract these rules from historical data is crucial for predicting future trends. However, since these rules are often disturbed by external noise, noise reduction while preserving critical internal information is necessary to improve the accuracy of fuzzy time series forecasting. In this paper, we propose a novel two-factor high-order fuzzy time series (FTS) forecasting model based on hesitant probabilistic fuzzy logical relationship (HPLR). To evaluate the performance of the model, we conduct empirical analysis using the closing price of the Taiwan Stock Exchange Capitalization Weighted Stock Index (TAIEX) as the main factor and the opening price as the secondary factor. The proposed model shows improved prediction performance and is intelligent and interpretable in model design. In addition, we forecasted the Hang Seng Index (HSI) to further illustrate the generalizability of the model.

A Comparative Study on Fuzzy Time Series Forecasting and Autoregressive Integrated Moving Average Models

To ensure the sustainable growth of coal production, it is necessary to analyse the growth, to ground the plans and management decisions on effective diagnostics and prediction of current and future situation at the coal production. This study presents a application of fuzzy time series forecasting methods, The new technique is applied to forecasting the coal production data using a fuzzy approach. For testing the methodology, statistical data on the coal production from 1980- to 2019. The Sturges rule is proposed to use as the universe of discourse. The intervals of variation of such indicators as growth rate are calculated when applying the approach to all defined fuzzy sets. The ARIMA model algorithm was applied by using the R - software to find the forecasted values. The forecasting results, obtained by the fuzzy time series method, are supposed to have more accuracy rate than time series model.

Hesitant fuzzy time series forecasting model of higher order based on one and two-factor aggregate logical relationship

Fuzzy time series forecasting is a methodology that incorporates fuzzy logic to handle uncertainty in time series data. It allows for more flexible and accurate predictions by considering the linguistic terms and membership functions associated with each data point, enabling a better representation of complex patterns and trends within the time series. A time series forecasting model examines the relationships between present and past observations for future measurement. This article proposes a hesitant fuzzy time series forecasting (FTSF) technique for higher order based on one and two-factor aggregate fuzzy relationships. The proposed method presents a new approach to partitioning intervals into an unequal and equal length. The hesitant fuzzy sets are defined for each observation, and fuzzification is based on their maximum score function. Further, aggregated triangular fuzzy numbers are defined using an aggregation operator corresponding to the fuzzified datum. In order to enhance forecasting performance, this work also introduces a novel defuzzification rule based on aggregated fuzzy logical relationship groups (AFLRGs) for higher order one and two-factor fuzzy time series. The implementation of the proposed approach is verified on two data sets to the forecasting of enrollment data and TAIFEX and TAIEX stock index data sets, and the forecasting error is tested in term of different statistical parameters. Thus, the model validates the effectiveness of the proposed method with higher forecasting accuracy rate in the environment hesitant fuzzy information.

Forecasting the Composite Stock Price Index Using Fuzzy Time Series Type 2

The movement and fluctuation of the IHSG are one of the references for investors in making investment decisions for buying, selling, or holding share ownership. Forecasting the value of the IHSG can assist investors in making this decision. This study used the fuzzy time series type 2 method to predict the IHSG. This study uses monthly IHSG data for 2017-2021 with 3 variables, namely close prices, high prices, and low prices. In fuzzy time series forecasting, the length of the interval affects the prediction results. This study uses a distribution and average-based method in determining the length of the interval to obtain optimal forecasting results. Based on the calculation of MAPE, the forecasting using average-based and distribution-based interval lengths had errors of 2.56% and 2.46%. The MAPE value shows that the forecasting results from the two methods of taking the length of the interval are very good. The IHSG forecasting results in January 2022 use a distribution-based interval length is 6450 while the IHSG forecasting results in January 2022 use an average-based interval length is 6510. The results of this study indicate that the interval length affects the forecasting results in fuzzy time series.
 
 Keywords: IHSG, fuzzy time series type 2, length of the interval, MAPE

Fuzzy Time Series Forecasting Accuracy Based on Hybrid Similarity Measure

The majority of fuzzy time series forecasting (FTSF) algorithms assess forecasting accuracy using an error-based distance. The predicted value is defuzzified to a crisp number and the error-based distance will be computed. Defuzzification causes some information to be lost, which leads to its inability to comprehend the level of uncertainty that has been preserved during the forecasting process. This paper proposes an enhanced FTSF model with forecasting accuracy developed based on a new hybrid similarity measure combining the centre of gravity and area and height. Three properties of the hybrid similarity measure are presented. The FTSF model is implemented in the case of the Malaysian unemployment rate. The findings indicate that, on average more than 94% of the predicted value was identical to historical data. The forecasting accuracy is produced directly from the forecasting value without undergoing the defuzzification process, which can preserve some information from being lost.

Kapılı Tekrarlayan Hücreler Tabanlı Bulanık Zaman Serileri Tahminleme Modeli

Time series forecasting has lots of applications in various industries such as weather, business, retail and energy consumption forecasting. Accurate prediction in these applications is very important and also difficult task because of complexity and uncertainty of time series. Nowadays, using deep learning methods is a popular approach in time series forecasting and shows better performance than classical methods. However, in the literature, there are few studies which use deep learning methods in fuzzy time series (FTS) forecasting. In this study, we propose a novel FTS forecasting model which is based upon hybridization of Recurrent Neural Networks with FTS to deal with complexity and also uncertanity of these series. The proposed model utilizes Gated Recurrent Unit (GRU) to make prediction by using combination of membership values and also past value from original time series data as model input and produce real forecast value. Moreover, the proposed model can handle first order fuzzy relations as well as high order ones. In experiments, we have compared our model results with those of state-of-art methods by using two real world datasets; The Taiwan Stock Exchange Capitalization Weighted Stock Index (TAIEX) and Nikkei Stock Average. The results indicate that our model outperforms or performs similar to other methods. The proposed model is also validated by using Covid-19 active case dataset and shows better performance than Long Short-term Memory (LSTM) networks.

Fuzzy Time Series Forecasting: Chen, Markov Chain and Cheng Models

Fuzzy Time Series Forecasting: Chen, Markov Chain and Cheng Models

A novel high order hesitant fuzzy time series forecasting by using mean aggregated membership value with support vector machine

Over the last three decades, several researchers have been putting their efforts into developing non-deterministic fuzzy time series (FTS) models using the traditional fuzzy set. However, considering a set of membership values to each element of the time series, the hesitant fuzzy set glorifies the chances to capture the fuzziness and uncertainty due to randomness better than the traditional fuzzy set. Motivated by this, the present study proposes a novel hesitant FTS forecasting model employing a support vector machine (HFTSF-SVM). In this model, to diminish the computational complexity and increase the forecasting accuracy, we have addressed three key issues by proposing three new methods such as (a) a length-based discretization technique is developed to compute the number of intervals, (b) two sets of universe of discourse are determined for the same time series to compute the non-membership and membership value for each observation (c) each observation along with its mean aggregated membership values are adopted to construct the fuzzy logical relationships (FLR). To assess the competency and reliability of the proposed HFTSF-SVM model, six promising FTS models from the recent literature and four promising time series forecasting (TSF) models are considered for comparison using sixteen time series datasets. Extensive statistical analyses of obtained results confirm the supremacy of the proposed HFTSF-SVM model.

Implementation of Revised Heuristic Knowledge in Average-based Interval for Fuzzy Time Series Forecasting of Tuberculosis Cases in Sabah

Implementation of Revised Heuristic Knowledge in Average-based Interval for Fuzzy Time Series Forecasting of Tuberculosis Cases in Sabah

Forecasting Palm Oil Production Using Fuzzy Time Forecasting Two-Factor Cross Associations with Frequency Density Partitions

Economic decisions have many determining factors based on estimates of macroeconomic variables. The accuracy of decision estimates can have an important impact. Forecasting is a method to reducing uncertainty about the future, because of economic decisions have multi-factor problems, the high order fuzzy time series forecast method is more suitable than the first order fuzzy time series forecast. Predictions are made for main factors by taking influence from both factors. FLR reflects the relationship between the premise and consequence. In this paper will be discussed fuzzy time series forecasting multi-factor one order cross association based on frequency density partition as a forecasting method to forecast palm oil production with influenced by large of the area. The results of the estimates show that the proposed method has a high forecast performance, with AFER value is according to the AFER criteria table 10%, it can be concluded that the forecast has very good criteria

Improving the Accuracy of Forecasting Models Using the Modified Model of Single-Valued Neutrosophic Hesitant Fuzzy Time Series

Proposed in this study is a modified model for a single-valued neutrosophic hesitant fuzzy time series forecasting of the time series data. The research aims at improving the previously presented single-valued neutrosophic hesitant fuzzy time series (SVNHFTS) model by including several degrees of hesitancy to increase forecasting accuracy. The Gaussian fuzzy number (GFN) and the bell-shaped fuzzy number (BSFN) were used to incorporate the degree of hesitancy. The cosine measure and the single-valued neutrosophic hesitant fuzzy weighted geometric (SVNHFWG) operator were applied to analyze the possibilities and pick the best one based on the neutrosophic value. Two data sets consist of the short and low-frequency time series data of student enrollment and the long and high-frequency data of ten major cryptocurrencies. The empirical result demonstrated that the proposed model provides higher efficiency and accuracy in forecasting the daily closing prices of ten major cryptocurrencies compared to the S-ANFIS, ARIMA, and LSTM methods and also outperforms other FTS methods in predicting the benchmark student enrollment dataset of the University of Alabama in terms of computation time, the Mean Absolute Percentage Error (MAPE), Mean Absolute Error (MAE), and the Root Mean Square Error (RMSE).

Improved seagull optimization algorithm of partition and XGBoost of prediction for fuzzy time series forecasting of COVID-19 daily confirmed

The establishment of fuzzy relations and the fuzzification of time series are the top priorities of the model for predicting fuzzy time series. A lot of literature studied these two aspects to ameliorate the capability of the forecasting model. In this paper, we proposed a new method(FTSOAX) to forecast fuzzy time series derived from the improved seagull optimization algorithm(ISOA) and XGBoost. For increasing the accurateness of the forecasting model in fuzzy time series, ISOA is applied to partition the domain of discourse to get more suitable intervals. We improved the seagull optimization algorithm(SOA) with the help of the Powell algorithm and a random curve action to make SOA have better convergence ability. Using XGBoost to forecast the change of fuzzy membership in order to overcome the disadvantage that fuzzy relation leads to low accuracy. We obtained daily confirmed COVID-19 cases in 7 countries as a dataset to demonstrate the performance of FTSOAX. The results show that FTSOAX is superior to other fuzzy forecasting models in the application of prediction of COVID-19 daily confirmed cases.

Fuzzy-Based Time Series Forecasting and Modelling: A Bibliometric Analysis

The purpose of this paper is to present the results of a systematic literature review regarding the development of fuzzy-based models for time series forecasting in the period 2017–2021. The study was conducted using a well-established review protocol and a couple of powerful tools for bibliometric analysis to know and analyse the main approaches adopted in the research field of interest. We analysed 118 articles published in peer-reviewed journals indexed in the 2020 Journal Citation Reports of the Web of Science. This allowed us to present an in-depth performance analysis and a science mapping regarding the current situation of fuzzy time series forecasting and modelling. The outputs of this study provide a practical base for further investigations that address this topic from both a methodological point of view and in terms of applicability.

Fuzzy Time Series Forecasting Approach using LSTM Model

In the present scenario, fuzzy time series forecasting (FTSF) is an interesting concept by the researchers to approach the uncertainty in the dataset. In the current study, we proposed a fuzzy long short term memory (FLSTM) model to forecast a wide range of time series (TS) dataset with less computational complexity. The present research mainly focuses on two issues such as 1) in order to obtain the number of intervals (NOIs) of the universe of discourse (UOD) the trend based discretization (TBD) approach is applied, and 2) the subscript of the fuzzy set associated with the crisp observation is considered to establish the fuzzy logical relationships (FLRs) for the proposed FLSTM model. To demonstrate the forecasting ability of the FLSTM model, six TS datasets with three profound FTSF models are considered. The empirical result analysis revealed that, in all measured the proposed model outperformed and showed better result than its alternatives. The outcome of the different FTSF models on different measures proves the outperformance of the FLSTM model than its competitors.

Using fuzzy time series forecasting and Gaussian mixture model to classify and predict new cases of COVID-19 in Saudi Arabia

In light of the global events resulting from the spread of the Corona pandemic and viral mutations, there is a need to examine epidemic data in terms of numbers of infected and deaths, different geographical locations, and the dynamics of disease dissemination virus. In the Kingdom of Saudi Arabia (KSA), since the spread of the virus on March 2, 2020, the number of confirmed cases has increased to 599044 cases until January 13, 2022, of which 262 are critical cases, while the number of recovery cases have reached 55035 cases, and deaths are 8901. It is a serious disease, and its spread is difficult to contain. The number of cases has continued to grow rapidly since the first cases appeared. Guess and Buck?s model for forecasting time-series data is an important figure that cannot be crossed when predicting fuzzy time-series, although several modifications have been made to the model to improve the accuracy of its results. The Gaussian mixture model and the fuzzy method for modelling new cases in Saudi Arabia were used as machine learning methods to classify and predict new cases of the virus in Saudi Arabia. Foggy time series forecasting. The studied datasets from the World Health Organization from May 15 to August 12, 2020 were used.