Search

This paper proposes a new model for creating fuzzy time series and conducting future forecasts using the Bi-directional Long Short-term Memory model (Bi-LSTM). This study contributes by using both classical machine learning and deep learning techniques. Firstly, a self-updating clustering algorithm is used to determine the optimal number of clusters for time series. Secondly, the study describes a method for determining the fuzzy relationship between elements of time series and the identified clusters. Thirdly, the developed model creates the fuzzy time series based on the established rules. Finally, forecasting for the future value using the Bi-LSTM model with some improvement in optimizing the parameters and increasing the accuracy of forecasting result. The steps of the proposed model are clearly illustrated by a numerical example. Moreover, this model also tested on various datasets, including M3, M4, and M5 datasets. By evaluating its performance using parameters such as sMAPE and RMSSE, the proposed model has shown significant enhancements when compared to existing models. Additionally, the model's effectiveness is further demonstrated through its successful application in forecasting the VN index stock in Vietnam. [ABSTRACT FROM AUTHOR]

Abstract: A belief rule base inference methodology using the evidential reasoning (RIMER) approach has been developed recently. A belief rule base (BRB), which can be treated as a more generalized expert system, extends traditional IF-THEN rules, but requires the assignment of some system parameters including rule weights, attribute weights, and belief degrees. These parameters need to be determined with care for reliable system simulation and prediction. Some off-line optimization models have been proposed, but it is expensive to train and re-train these models in particular for large-scale systems. Moreover, the recursive algorithms are also proposed to fine tune a BRB online, which require less calculation time and satisfy the real-time requirement. However, the earlier mentioned learning algorithms are all based on a predetermined structure of the BRB. For a complex system, prior knowledge may not be perfect, which leads to the construction of an incomplete or even inappropriate initial BRB structure. Also, too many rules in an initial BRB may lead to over fitting, whilst too few rules may result in under fitting. Consequently, such a BRB system may not be capable of achieving overall optimal performance. In this paper, we consider one realistic and important case where both a preliminary BRB structure and system parameters assigned to given rules can be adjusted online. Based on the definition of a new statistical utility for a belief rule as investigated in this paper, a sequential learning algorithm for online constructing more compact BRB systems is proposed. Compared with the other learning algorithms, a belief rule can be automatically added into the BRB or pruned from the BRB, and our algorithm can also satisfy the real-time requirement. In addition, our algorithm inherits the feature of RIMER, i.e., only partial input and output information is required, which could be either incomplete or vague, either numerical or judgmental, or mixed. In order to verify the effectiveness of the proposed algorithm, a practical case study about oil pipeline leak detection is studied and examined to demonstrate how the algorithm can be implemented. [Copyright &y& Elsevier]

• The proposal of trend detection method from a stream of tweets. • Precise prediction & detection of real-time Twitter trends along with terms ranking. • Discusses applications of proposed research in business, industries and advertisers. Social media has recently become popular due to its vast applications. The common people all over the world uses its diverse channels to express personal views, experiences and opinions regarding diverse topics. Social media has revolutionized the way people interact and communicate with each other and overall, it has changed the methods and approaches in about all the aspects of life such as social issue, business, education, health, etc. Thus, sales and marketing departments of multinational industries are focusing on social media trends to analyze current trends and predict future trends by analyzing user generated content on Facebook, Flickr, Twitter, etc. However, the prediction process becomes challenging as the multiplicity of factors affect the popular elements in the social media content. This research paper aims to work on Twitter trend analysis and proposes a trend detection process over streams of tweets. The proposed approach detects the trending topics of the real-time Twitter trends along with ranking the top terms and hashtags. The paper further discusses the motivation for trend prediction over the social media; In addition to exploratory data analysis, the research paper explores the Term Frequency-Inverse Document Frequency (Tf-IDF), Combined Component Approach (CCA) and Biterm Topic Model (BTM) approaches for finding the topics and terms within given topics. In modern competitive world, this research provides investors, advertisers, industries and all the stakeholders. A detailed and comprehensive data analysis which may help them to focus their investment, area of work, marketing, and product. [ABSTRACT FROM AUTHOR]

By present, many models of bankruptcy forecasting have been developed, but this area remains a field of research activity; little is known about the practical application of existing models. In our opinion, this is because the use of existing models is limited by the conditions in which they are developed. Another question concerns the factors that can be significant for forecasting. Many authors suggest that indicators of the external environment, corporate governance as well as firm size contain important information; on the other hand, the large number of factors does not necessary increase predictive ability of a model. In this paper, we suggest the genetic algorithm based two-step classification method (TSCM) that allows both selecting the relevant factors and adapting the model itself to application. Classifiers of various models are trained at the first step and combined into the voting ensemble at the second step. The combination of random sampling and feature selection techniques were used to ensure the necessary diversity level of classifiers at the first step. The genetic algorithms are applied at the step of features selection and then at the step of weights determination in ensemble. The characteristics of the proposed method have been tested on the balanced set of data. It included 912 observations of Russian companies (456 bankrupts and 456 successful) and 55 features (financial ratios and macro/micro business environment factors). The proposed method has shown the best accuracy (0.934) value among tested models. It has also shown the most balanced precision-recall ratio. It found bankrupts (recall = 0.953) and not bankrupts (precision = 0.910) rather accurately than other tested models. The ability of method to select the task-relevant features has been also tested. Excluding the features that are significant for less than 50% of the classifiers in the ensemble improved the all performance metrics (accuracy = 0.951, precision = 0.932, recall = 0.965). So, the proposed method allows to improve the advantages and alleviate the weaknesses inherent in ordinary classifiers, enabling the business decisions support with a higher reliability. [ABSTRACT FROM AUTHOR]

Accurate and reliable forecasting is a crucial task in many different domains. The selection of a forecasting algorithm that is suitable for a specific time series can be a challenging task, since the algorithms' performance depends on the time-series properties, as well as the properties of the forecasting algorithms. The methodology and analysis presented in this paper are contributing towards understanding the performance of time-series forecasting methods. Instead of using time-series meta-features only to obtain a good meta-model that can predict the performance of a forecasting algorithm, the methodology can link which features are important for which forecasting methods. We used time-series meta-features extracted using the tsfresh and catch22 libraries. We also found that the importance of the meta-features changes depending on the meta-model that is used. There are only a few meta-features that always appear important for a given forecasting method no matter which meta-model will be used for learning, which further provides opportunities to select a model-agnostic feature portfolio. In addition, different feature importance techniques can provide different results that are related to the methodology that is used by the meta-model. By using the feature importance obtained by a meta-model and a specified feature importance technique, we can define a representation of a forecasting method behavior, which can further provide an insight into which forecasting methods have similar behavior. • Meta-learning performance prediction for time-series forecasting algorithms. • Analysis of features in the catch22 and tsfresh libraries. • Time-series feature importance analysis for algorithm performance prediction. [ABSTRACT FROM AUTHOR]

Accurate and reliable forecasting is a crucial task in many different domains. The selection of a forecasting algorithm that is suitable for a specific time series can be a challenging task, since the algorithms' performance depends on the time-series properties, as well as the properties of the forecasting algorithms. The methodology and analysis presented in this paper are contributing towards understanding the performance of time-series forecasting methods. Instead of using time-series meta-features only to obtain a good meta-model that can predict the performance of a forecasting algorithm, the methodology can link which features are important for which forecasting methods. We used time-series meta-features extracted using the tsfresh and catch22 libraries. We also found that the importance of the meta-features changes depending on the meta-model that is used. There are only a few meta-features that always appear important for a given forecasting method no matter which meta-model will be used for learning, which further provides opportunities to select a model-agnostic feature portfolio. In addition, different feature importance techniques can provide different results that are related to the methodology that is used by the meta-model. By using the feature importance obtained by a meta-model and a specified feature importance technique, we can define a representation of a forecasting method behavior, which can further provide an insight into which forecasting methods have similar behavior. • Meta-learning performance prediction for time-series forecasting algorithms. • Analysis of features in the catch22 and tsfresh libraries. • Time-series feature importance analysis for algorithm performance prediction. [ABSTRACT FROM AUTHOR]

Link prediction refers to predicting the connection probability between two nodes in terms of existing observable network information, such as network structural topology and node properties. Although traditional similarity-based methods are simple and efficient, their generalization performance varies widely in different networks. In this paper, we propose a novel link prediction approach ICP based on inductive matrix completion, which recoveries node connection probability matrix by applying node features to a low-rank matrix. The approach first explores a comprehensive node feature representation by combining different structural topology information with node importance properties via feature construction and selection. The selected node features are then used as the input of a supervised learning task for solving the low-rank matrix. The node connection probability matrix is finally recovered by a bi-linear function, which predicts the connection probability between two nodes with their features and the low-rank matrix. In order to demonstrate the ICP superiority, we took eleven related efforts including two recent methods proposed in 2020 as baseline methods, and it is shown that ICP has stable performance and good universality in twelve different real networks. Compared with the baseline methods, the improvements of ICP in terms of the average AUC results are ranging from 3.81% ∼ 12.77% and its AUC performance is improved by 0.08% ∼ 3.54% compared with the best baseline method. The limitation of ICP lies in its high computational complexity due to the feature construction, but the complexity can be reduced by replacing complex features with node semantic attributes if there are additional data available. Moreover, it provides a potential link prediction solution for large-scale networks, since inductive matrix completion is a supervised learning task, in which the underlying low-rank matrix can be solved by representative nodes instead of all their nodes. • A novel link prediction approach based on inductive matrix completion is proposed. • A comprehensive node feature representation is employed. • Link prediction is based on the node features and the factorized matrices. • Comparisons have been done with both classical and state-of-the-art methods. • The approach provides a potential link prediction solution for large-scale networks. [ABSTRACT FROM AUTHOR]

Abstract: In general, times series forecasting is considered as a highly complex problem, which is particularly true for financial time series. In this paper, a fuzzy model evolved through a bio-inspired algorithm is proposed to produce accurate models for the prediction of these time series. The performance of this model is compared to that of a group of state-of-the-art statistical models. A thorough experimental study is designed and carry out in order to assess the merits of the proposal. The experimental results allow us to state that our proposal forecasts consistently outperform the other considered methods. [Copyright &y& Elsevier]

Abstract: This study presents an adaptive intelligent algorithm for forecasting gasoline demand based of artificial neural network (ANN), conventional regression and design of experiment (DOE). To show the superiority and applicability of the proposed algorithm it has been applied for monthly demand estimation of gasoline in Japan, USA, Kuwait, Canada and Iran within 1992–2005. The economic indicators used in this paper are price, GDP (Gross Domestic Production), population, number of vehicles, gasoline demand in the last periods and correlation coefficient between variables. The proposed algorithm may be used to estimate gasoline demand in the future by optimizing parameter values. The proposed algorithm uses ANOVA to select either ANN or regression for future demand estimation. Furthermore, if the null hypothesis in ANOVA and F-test is rejected, the Duncan method is used to identify which model is closer to actual data at α level of significance. Neural network is an intelligent, adaptive and optimizing method for solving the complicated problems. The results show that ANN provides far less error than regression. Moreover, the proposed algorithm provides near optimal solution for gasoline demand estimation due to its flexibility and intelligibility. The intelligent algorithm is therefore ideal for policy making with respect to gasoline consumption estimation. [ABSTRACT FROM AUTHOR]

Abstract: This study develops an improved fuzzy time series method via adjustment of the latest value factor and previous error patterns. There are many fuzzy extended applications in the literature, and the fuzzy time series is one successful implementation of fuzzy logical modelling. Fuzzy time series have been studied for over a decade, and many researchers have proposed to remove some of the drawbacks of the initial fuzzy time series algorithm. In this paper, fuzzy integrated logical forecasting (FILF) and extended FILF (E-FILF) algorithms are suggested for short term forecasting purposes. Empirical studies are performed over the Baltic Dry Index (BDI), and indicate the superiority of the proposed approach compared to conventional benchmark methods. [Copyright &y& Elsevier]

Abstract: Voting features based classifiers, shortly VFC, have been shown to perform well on most real-world data sets. They are robust to irrelevant features and missing feature values. In this paper, we introduce an extension to VFC, called voting features based classifier with feature construction, VFCC for short, and show its application to the problem of predicting if a bank will encounter financial distress, by analyzing current financial statements. The previously developed VFC learn a set of rules that contain a single condition based on a single feature in their antecedent. The VFCC algorithm proposed in this work, on the other hand, constructs rules whose antecedents may contain conjuncts based on several features. Experimental results on recent financial ratios of banks in Turkey show that the VFCC algorithm achieves better accuracy than other well-known rule learning classification algorithms. [Copyright &y& Elsevier]

In recent years, some researchers focused on the research topic of using fuzzy time series to handle forecasting problems. In this paper, we present a new method to forecast enrollments based on automatic clustering techniques and fuzzy logical relationships. First, we present an automatic clustering algorithm for clustering historical enrollments into intervals of different lengths. Then, each obtained interval will be divided into p sub-intervals, where . Based on the new obtained intervals and fuzzy logical relationships, we present a new method for forecasting the enrollments of the University of Alabama. The proposed method gets a higher average forecasting accuracy rate than the existing methods. [Copyright &y& Elsevier]

Abstract: Many forecasting models based on the concept of fuzzy time series have been proposed in the past decades. Two main factors, which are the lengths of intervals and the content of forecast rules, impact the forecasted accuracy of the models. How to find the proper content of the main factors to improve the forecasted accuracy has become an interesting research topic. Some forecasting models, which combined heuristic methods or evolutionary algorithms (such as genetic algorithms and simulated annealing) with the fuzzy time series, have been proposed but their results are not satisfied. In this paper, we use the particle swarm optimization to find the proper content of the main factors. A new hybrid forecasting model which combined particle swarm optimization with fuzzy time series is proposed to improve the forecasted accuracy. The experimental results of forecasting enrollments of students of the University of Alabama show that the new model is better than any existing models, and it can get better quality solutions based on the first-order and the high-order fuzzy time series, respectively. [Copyright &y& Elsevier]

Noisy time series prediction is attractive and challenging since it is essential in many fields, such as forecasting, modeling, signal processing, economic and business planning. Radial basis function (RBF) neural network is considered as a good candidate for the prediction problems due to its rapid learning capacity and, therefore, has been applied successfully to nonlinear time series modeling and forecasts. However, the traditional RBF network encounters two primary problems. The first one is that the network performance is very likely to be affected by noise. The second problem is about the determination of the number of hidden nodes. In this paper, we present an M-estimator based robust radial basis function (RBF) learning algorithm with growing and pruning techniques. The Welsch M-estimator and median scale estimator are employed to get rid of the influence from the noise. The concept of neuron significance is adopted to implement the growing and pruning techniques of network nodes. The proposed method not only eliminates the influence of noise, but also dynamically adjusts the number of neurons to approach an appropriate size of the network. The results from the experiments show that the proposed method can produce a minimum prediction error compared with other methods. Furthermore, even adding 30% additive noise of the magnitude of the data, this proposed method still can do a good performance. [Copyright &y& Elsevier]

In this paper, we present a mining algorithm to improve the efficiency of finding large itemsets. Based on the concept of prediction proposed in the (n, p) algorithm, our method considers the data dependency in the given transactions to predict promising and non-promising candidate itemsets. Our method estimates for each level a different support threshold that is derived from a data dependency parameter and determines whether an item should be included in a promising candidate itemset directly. In this way, we maintain the efficiency of finding large itemsets by reducing the number of scanning the input dataset and the number candidate items. Experimental results show our method has a better efficiency than the a priori and the (n, p) algorithms when the minimum support value is small. [Copyright &y& Elsevier]

Abstract: In this paper we propose and implement a fusion model by combining the Hidden Markov Model (HMM), Artificial Neural Networks (ANN) and Genetic Algorithms (GA) to forecast financial market behaviour. The developed tool can be used for in depth analysis of the stock market. Using ANN, the daily stock prices are transformed to independent sets of values that become input to HMM. We draw on GA to optimize the initial parameters of HMM. The trained HMM is used to identify and locate similar patterns in the historical data. The price differences between the matched days and the respective next day are calculated. Finally, a weighted average of the price differences of similar patterns is obtained to prepare a forecast for the required next day. Forecasts are obtained for a number of securities in the IT sector and are compared with a conventional forecast method. [Copyright &y& Elsevier]