Usually, machine and automated guided vehicle (AGV) scheduling are studied simultaneously. However, previous studies often used a fixed number of AGVs or did not consider routing problems transportation time. This paper focuses on the AGV problem in flexible manufacturing system by simultaneously considering optimal AGVs, shortest time, path planning problem, conflict-free (CFRP). To study these simultaneously, we propose genetic algorithm combined with Dijkstra that is based time window....

Abstract As transport sector takes a big share of the whole energy consumption in China, it is crucial to predict its demand. To forecast China's demand, group method data handling ( GMDH ) was introduced. The model can help policymakers’ select influential variables and build prediction models automatically. Furthermore, reduce negative impact noise Chinese statistical data. produce comparable results, four six sets used this paper contain same as previously published research. Artificial...

Customer churn prediction is an important problem in customer relationship management. Experimental results from a novel multiple classifiers ensemble selection model based on the group method of data handling (GMDH) are encouraging.

Dynamic classifier selection (DCS) plays a strategic role in the field of multiple classifiers system. This article introduces group method data handing (GMDH) theory to DCS, and presents novel strategy GAES for adaptive ensemble first. Then it extends this algorithm proposes dynamic based on GMDH (GDES). For each test pattern, GDES is able select an appropriate from pool adaptively, determine combination weights among base classifiers, complete process automatically. We experimentally over...

Most of the existing models for oil price forecasting only use data in forecasted time series itself. This study proposes a transfer learning based analog complexing model (TLAC). It first transfers some related source domain to assist modeling target by technique, and then constructs method. Finally, genetic algorithm is introduced find optimal matching between two important parameters TLAC. Two main crude series, West Texas Intermediate (WTI) spot Brent are used empirical analysis, results...

Abstract Based on the theories and methods of self‐organizing data mining, a new forecasting method, called combining is proposed. Compared with optimal linear neural networks methods, method can improve capability model. The superiority justified demonstrated by real applications. Copyright © 2005 John Wiley & Sons, Ltd.

Emergency department (ED) crowding has developed into a risk that severely impacts on the emergency medical service quality and access to health care. With regard essential cause of crowding, subjectively, it results from low efficiency resource collocation or incomprehension patient flow rule; objectively, arises rate. Hence, prediction in is significant providing early warning ED for interest staff, monitoring controlling condition policy alleviate crowding. The proposed methods...

While biometric fusion is a well-studied problem, most of schemes cannot account for missing data (incomplete score lists), that commonly encountered in large-scale multibiometric identification systems. In this paper, we present new approach, where the RIBG (Robust Imputation Based on Group method handling) used handling data. Since scheme can be followed by standard designed complete data, propose Bees Algorithm based Weighted Sum Method (BASM) to find optimal parameters fuse information...

Mobile banking is regarded as main recommended business of all commercial banks at present.Its research in academia very deficient.In the paper, influence relationship perceived value on loyalty discussed.The customer mobile subdivided into functional value, service emotional social and safety according to study theory.Hypothetical model with satisfaction mediating variable proposed.Method reliability validity analysis, correlation analysis regression sample SPSS utilized for verifying...

The purpose of this study is to find the key factors amount outstanding balances among revolving credit card users in Chinese market. A Heckman procedure used analyze a dataset bank China. small revolver coursing imbalanced problem, and we try use rebalanced method machine learning domain deal with problem. Results show there are differences determinants being balance. Age, housing condition, industry, average cash advance per time, etc. significant related

Clustering is popular used in customer value segmentation business research. Compared with other clustering methods, the objective analysis can automatically and objectively determine number of clusters find out optimal scheme. This investigation discussed reasonable evaluation system value-driven segmentation, identified behavior using a recency, frequency monetary(RFM) index basic properties as integrated variables, then, presented novel approach-objective that be value-based segmentation....

Scientific customer value segmentation (CVS) is the base of efficient relationship management, and credit scoring, fraud detection, churn prediction all belong to CVS. In real CVS, data usually include lots missing values, which may affect performance CVS model greatly. This study proposes a one-step dynamic classifier ensemble for values (ODCEM) model. On one hand, ODCEM integrates preprocess classification modeling into step; on other it utilizes multiple classifiers technology in...