A rainfall forecasting method using machine learning models and its application to the Fukuoka city case

Sumi, Sirajum Monira; Zaman, M. Faisal; Hirose, Hideo

doi:10.2478/v10006-012-0062-1

Cited by 88 publications

(30 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Artificial neural network methodologies and their combinations like fuzzy neural networks are some of the well-known methods of forecasting, which are successfully utilized and evaluated in various fields including supply chain (see the works of Efendigil et al (2009) (in fuzzy), Brdyś et al (2009) (in stock exchange), Sumi et al (2012) (in rainfall forecasting), Ozkr and Balgil (2013) (in the fuzzy approach), Georgiadis (2013) (in system dynamics), and Soleimani et al (2014) (in risk management)). Indeed, the ability of the self-learning of the neural network-based methodologies makes them powerful techniques of forecasting.…”

Section: Introductionmentioning

confidence: 99%

Forecasting return products in an integrated forward/reverse supply chain utilizing an ANFIS

Kumar¹,

Soleimani

Govindan

2014

International Journal of Applied Mathematics and Computer Science

View full text Add to dashboard Cite

Interests in Closed-Loop Supply Chain (CLSC) issues are growing day by day within the academia, companies, and customers. Many papers discuss profitability or cost reduction impacts of remanufacturing, but a very important point is almost missing. Indeed, there is no guarantee about the amounts of return products even if we know a lot about demands of first products. This uncertainty is due to reasons such as companies' capabilities in collecting End-of-Life (EOL) products, customers' interests in returning (and current incentives), and other independent collectors. The aim of this paper is to deal with the important gap of the uncertainties of return products. Therefore, we discuss the forecasting method of return products which have their own open-loop supply chain. We develop an integrated two-phase methodology to cope with the closed-loop supply chain design and planning problem. In the first phase, an Adaptive Network Based Fuzzy Inference System (ANFIS) is presented to handle the uncertainties of the amounts of return product and to determine the forecasted return rates. In the second phase, and based on the results of the first one, the proposed multi-echelon, multi-product, multi-period, closed-loop supply chain network is optimized. The second-phase optimization is undertaken based on using general exact solvers in order to achieve the global optimum. Finally, the performance of the proposed forecasting method is evaluated in 25 periods using a numerical example, which contains a pattern in the returning of products. The results reveal acceptable performance of the proposed two-phase optimization method. Based on them, such forecasting approaches can be applied to real-case CLSC problems in order to achieve more reliable design and planning of the network.Keywords: artificial neural network, adaptive network based fuzzy inference system, closed-loop supply chain, forecasting methods, fuzzy neural network.

show abstract

Section: Introductionmentioning

confidence: 99%

Forecasting return products in an integrated forward/reverse supply chain utilizing an ANFIS

Kumar¹,

Soleimani

Govindan

2014

International Journal of Applied Mathematics and Computer Science

View full text Add to dashboard Cite

show abstract

“…More precisely, performing a PCA includes the computation of a static eigenbasis, usually with the aim of exploratory data analysis (Lenz and Bui, 2004;Turk and Pentland, 1991). A typical task is to project various vectors of data using an identical eigenbasis, in order to obtain a representation that best explains the variance in the data (Han, 2010;Sumi et al, 2012). While PCA is continuously applied in various fields for statistical analysis, such as recently in the work of Skraban et al (2013), it has also been successfully adapted for solving more complex tasks, e.g., for load forecasting in power systems (Siwek et al, 2009) and for automated recognition of faces (Liu et al, 2003;Turk and Pentland, 1991).…”

Section: Introductionmentioning

confidence: 99%

An efficient eigenspace updating scheme for high-dimensional systems

Gangl¹,

Mongus²,

Žalik³

2014

International Journal of Applied Mathematics and Computer Science

View full text Add to dashboard Cite

Systems based on principal component analysis have developed from exploratory data analysis in the past to current data processing applications which encode and decode vectors of data using a changing projection space (eigenspace). Linear systems, which need to be solved to obtain a constantly updated eigenspace, have increased significantly in their dimensions during this evolution. The basic scheme used for updating the eigenspace, however, has remained basically the same: (re)computing the eigenspace whenever the error exceeds a predefined threshold. In this paper we propose a computationally efficient eigenspace updating scheme, which specifically supports high-dimensional systems from any domain. The key principle is a prior selection of the vectors used to update the eigenspace in combination with an optimized eigenspace computation. The presented theoretical analysis proves the superior reconstruction capability of the introduced scheme, and further provides an estimate of the achievable compression ratios.

show abstract

“…To build a numerical predictive model of any process it is necessary to define the set of input features (also called explanatory variables) on the basis of which the forecasting will be made (Sumi et al, 2012). This choice is made based a detailed analysis of the problem.…”

Section: Potential Set Of Featuresmentioning

confidence: 99%

Data mining methods for prediction of air pollution

Siwek

Osowski

2016

International Journal of Applied Mathematics and Computer Science

View full text Add to dashboard Cite

The paper discusses methods of data mining for prediction of air pollution. Two tasks in such a problem are important: generation and selection of the prognostic features, and the final prognostic system of the pollution for the next day. An advanced set of features, created on the basis of the atmospheric parameters, is proposed. This set is subject to analysis and selection of the most important features from the prediction point of view. Two methods of feature selection are compared. One applies a genetic algorithm (a global approach), and the other-a linear method of stepwise fit (a locally optimized approach). On the basis of such analysis, two sets of the most predictive features are selected. These sets take part in prediction of the atmospheric pollutants PM10, SO2, NO2 and O3. Two approaches to prediction are compared. In the first one, the features selected are directly applied to the random forest (RF), which forms an ensemble of decision trees. In the second case, intermediate predictors built on the basis of neural networks (the multilayer perceptron, the radial basis function and the support vector machine) are used. They create an ensemble integrated into the final prognosis. The paper shows that preselection of the most important features, cooperating with an ensemble of predictors, allows increasing the forecasting accuracy of atmospheric pollution in a significant way.

show abstract

A rainfall forecasting method using machine learning models and its application to the Fukuoka city case

Cited by 88 publications

References 41 publications

Forecasting return products in an integrated forward/reverse supply chain utilizing an ANFIS

Forecasting return products in an integrated forward/reverse supply chain utilizing an ANFIS

An efficient eigenspace updating scheme for high-dimensional systems

Data mining methods for prediction of air pollution

Contact Info

Product

Resources

About