AN IMPROVED CREDIT SCORING METHOD FOR CHINESE COMMERCIAL BANKS
|
|
- Sharyl Arnold
- 8 years ago
- Views:
Transcription
1 AN IMPROVED CREDIT SCORING METHOD FOR CHINESE COMMERCIAL BANKS Jianping Li Jinli Liu Weixuan Xu 1.University of Science & Technology of China, Hefei, , P.R. China 2.Institute of Policy and Management Chinese Academy of Sciences, Beijing , P.R. China 1.University of Science & Technology of China, Hefei, , P.R. China 2.Institute of Policy and Management Chinese Academy of Sciences, Beijing , P.R. China Institute of Policy and Management Chinese Academy of Sciences, Beijing , P.R. China Yong Shi College of information Science and Technology University of Nebraska at Omaha, Omaha, NE 68182, USA Abstract This paper presents an improved credit scoring to the Chinese commercial bank for credit card risk management. To get a comprehensive scoring for classifying the grade, the principal component analysis is used, which enables us to calculate the weights of the original indexes, and to get a comprehensive function for computing the score of new applicants. Comparing with other methods, this method has an outstanding merit that it gives objective weights of original indexes and can easily adapt to the different economic & cultural environments in different regions and the population drift in same region which cause the disparity of credit scoring. A demonstration test shows that the model complies with the practice in credit card risk management, hence has an application foreground. A comparison study indicates that this method has better results than those scoring methods currently used in banks. Keywords: credit card; risk management; credit scoring; principal component analysis 1. Introduction With the conception of 2003, the prime year of credit card in China put forward by the China Merchants Bank, the competition in credit card market is increasingly heated, and the scale of issuing credit card is expanding rapidly in China. As to the credit card risk management, how to analyze the customers credit risk effectively and establish the reasonable credit issue criteria has become a key task of the credit card issuers. In credit card risk management, credit scoring is widely used. For example, more than 97% of banks use the scoring system to decide the credit card loan s application in USA. Credit scoring is a quantitative method to deal with large quantities of small loans in commercial banks. Based on the scoring models established by the mass historical data in bank, credit scoring predicts the default rate of the loan applicants or the exited loan. After collecting the applicant s
2 information, the bank gives a score via the credit scoring system, and decides whether grant the loan or not in a very short time according to the score. At present, the Chinese banks also use the scoring method in card issuing as the assistant technique. The general process includes the following steps: Choose some characters and attach different weights to according classes, score the applicant according to his/her characters and decide whether or not grant credit and how much will be granted considering the applicant s total score to the credit grade. At present, the social credit system has not been fully established and takes little effect in Chinese commercial banks. Many reasons pose great barrier to applying the scoring method to practice, such as: (1) Lack of characters The popular credit scores such as FICO score and Credit Bureau score choose at least 50 to 60 variables, while the Chinese banks use only 15 characters or so. This leads to inadequate assessment of the applicant s risk. (2) The methods in credit scoring have adaptability problem in practice Amongst different regions of China, there are remarkable differences in the economic development level, ideas and other aspects. Therefore, using the same score criteria will inevitably cause the evaluating error. Even in the same region, as China is experiencing a high speed development in economy, society and culture, the economy environment, population structure and the life style are changing very fast, which is called the fast population drift. This also causes unconformity between the scoring result under original criteria and the real situation. It is required that the scoring model adjust the weights of the characters, the score of the detailed class and the credit grade scores according to different samples. The prevailing score model in Chinese commercial banks is far from this requirement. The first problem lies mainly in the influence of the external credit environment, and the improvement is also based upon the total credit environment reform, which is a stepwise process. This paper aims to solve the second problem the adaptability of score model. An improved method credit scoring is presented which has a preferable adaptability in theory. The demonstration shows it meets the need of credit risk management, and is superior to the current method in use. And the test results show it has a promising applicative foreground. 2. The basic thoughts and construction process of the improved method We construct a new credit score method trying to overcome the current method s problem of adaptability. Its basic thought is as follows: Credit scoring can be expressed by a kind of comprehensive function of each primitive index, y = f ( x1, x2,..., x n ). n is the number of primitive index. In our method, suppose y is a linear function. This linear comprehensive function could be obtained by analyzing the historical samples. And the applicants score can be easily calculated using this function. According to the distribution of the bad person during the score, the bank will set up different grades and establish the scoring criteria based on the concrete strategy and the need of business. Then, the bank use the comprehensive function to predict the new applicants score, and determine whether to grant them credit or not and how much credit to be granted according to their scores and the existed scoring criteria. In the comprehensive function, how to get the index coefficients is a key problem. These coefficients satisfy at least the following conditions: (1) Can be changed easily if the sample is changed; (2) Are objective; (3) And can guarantee the rationality of the assessment result We use the principal component analysis method to get the coefficients according to the above terms. Principal Components Analysis (PCA) is a multivariate statistical method to reduce the dimensions through transforming multiple indexes into fewer comprehensive indexes, which was put forward by Hotelling in Dimension-decline and giving the relative importance by the difference of data are the main characteristics of this method. In current Chinese personal credit assessment, there are few indexes for using, therefore, choosing how much principal factors is not the main problem under the current circumstances. Our target of using the PCA is not to emphasize its dimension-decline function to reduce the complicated
3 degree of the problem, but to concentrate on the following purposes: (1) To obtain the objective coefficients The PCA method emphasizes the difference principle, and creates the coefficients fully based on data themselves. So, the coefficients have the objectivity and avoid too many human s interferences. (2) To remove the assessment deviations caused by macroscopic factors and the population drift. According to the principle of PCA, we can get different principal factors and indexes from different samples. So, we can divide China into different regions according to economic environment and apply different samples to compute and induce different scoring functions. Using the different score and the bad persons distributions, we can set up the different credit criteria in different regions. In the same way, this method can also easily realize the dynamic change of the comprehensive function in the same region. Once the bank considers there is a big population drift in one region, it can adjust the comprehensive scoring function of the new sample. The detailed constructing process can be divided into the following steps: (1) Sample selection Select a sample that is composed of a mount of good and bad records in history. The good and the bad will be defined by each bank themselves. Divide the total sample into two sets stochastically: one is a training set to compute the comprehensive function; and the other is a test sample to verify the validity of the comprehensive function. (2) Data pretreatment In order to guarantee the accuracy of using the PCA method, we carry out the data pretreatment primarily to make the data have the same direction and to guarantee the indexes have economic meanings. In the light of this thought, we classify the sample data by the current division standard of one Chinese commercial bank. And then, take the odds of good/bad as the input data to PCA computation in every certain class. The bigger the data are, the better they will be. For example, the age index is divided into 5 concrete categories: 18-22,23-34,35-40,41-60 and > 60. If there are 60 good and 34 bad in the category, then the input data is 1.765(60/34). The other index may be handled similarly. (3) PCA calculation After the data pretreatment, the PCA could be calculated. Firstly, examination of PCA method s fitness is necessary. The tool for the examination is a KMO and Bartlett s spheroid test. If the KMO value is larger than 0.5, the Bartlett test is significant. This means the data are suitable for the PCA method. (4) To build the predicting function Suppose we have selected s principal factors, the scoring function of the principal factors in the training sample could be expressed as: rate : F = a1z1 + a2z as zs (1) zi is the i th principal factor, a i is its contribution a i λ λ i m = k m= 1 a tl = k λ λ m m= 1 l m m= 1 ( i, l = 1, 2,..., k ) th ( λ i is the i eigenvalue sorted by descending, a tl is the cumulative variance contributions of principal factor z to z ). Then, 1 l Z = bx+ bx+ + bx (2) i i1 1 i in n Take the equation (2) into equation (1), we get: F = c x + c x + + c X (3) c j n n s = a ib i j i = 1 ( i = 1, 2,..., s; j = 1, 2,..., n) (5) Setting up the credit grade According to the comprehensive score and the bad person s distribution in the training sample, we can set up the different grade by a certain method and standard. That is, one can classify the different class based on the comprehensive score. (6) Assessing new applicants After the same data pretreatment and data standardization like the train sample, the new applicants credit score could be predicted using equation (3). And it may be determined whether to grant him/her credit or not according to his/her score and the credit grade given by step 5. Figure 1 shows basic thought of the method and constructing process explicitly.
4 Original data Data pretreatment Principal Component Analysis Compute the comprehensive i Classify different credit grade according to the score and the odds of good/bad Compute the character weight Build the comprehensive function The score of new applicants Reject No Reach the cutoff? Yes Granting the corresponding credit Figure 1 the basic thought and construct process of our method varimax method rotation to conduct factor analysis, we 3. Demonstration results select ex-10 principal components, and the cumulative variance is 86.74%. Based on the method presented in part 2, here is a Table 1 KMO and Bartlett s test of spheroid demonstration of the model. We use a sample including Testing Value 1350 credit card records, and define two classes of credit card users: the bad and the good records were selected at random as the calculating sample to get the comprehensive function. These 1000 records include 755 KMO Measure of Sampling Adequacy good and 245 bad. The other records serve as the testing sample to test the validity of the comprehensive function. 262 good and 88 bad were within this sample. We choose 14 characters including age, income etc, and take the SPSS11.0 as the computing software. Bartlett's Test of phericity Approx.Chi-Square Df. 91 Sig The training results After the data pretreatment, SPSS gives KMO value and Bartlett examination, as is shown in the tale 1: The KMO value is 0.787>0.5 and the Bartlett examination is significant, so the principal component analysis is suitable. Take 0.8 as the cutoff to get the eigenvalue, using We use the equation 1 to compute the comprehensive score, and the result shows that the score of 1000-sample is very like a normal distribution, as the figure 2 shows: means is 0, and the standard deviation is (considering the actual condition, we magnify the score 100 times).
5 Std. Dev = Mean = 0.0 N = ZHUFEN Figure 2 the distributions of the sample scores According to this characteristic, we classify the score into 6 class based on the standard deviation, the means, one standard deviation and two standard deviation as the cutoffs. Table 2 shows the detail. Table 2 Training results in multiple-class Score 68 34~68 0~34-34~0-68~-34 <-68 Class Total Good credit Bad credit Total The bad rate Odds of good/bad *The score in every class includes the lower value and no t includes the upper value If the applicant s comprehensive score is larger than 68, this means larger than two standard deviation, accordingly be classified as the first grade and stands the highest credit. In this training sample, there are 28 records, 1 is bad and others are good, so the bad person s rate is 3.57% (1/28) in this grade. This bad rate can look like a type of default rate. Similarly, the comprehensive score between 34 to 68 belongs to the second grade, meaning a good credit in this class, and the bad person s ratio is 6.25%, and accordingly we get 6 grades. The bad person s ratio of every grade can be a measurement of credit risk in corresponding grade, the bigger the ratio is, the larger credit risk will be. According to the result, from the grade 1 to grade 6, with the reducibility of the comprehensive score, the default rate increases monotonously, that is, the lower comprehensive score, the bigger credit risk. This result accords with the theory of credit risk management. According to the general theories of the scoring, if the scoring of the applicant belongs to the grade that the odds of good/bad higher than 4, the bank will grant the credit to him/her. So, we classify the sample to two classes according to whether the odds of good/bad larger than 4, one class means good that will be granted the credit, and the other means bad that will be refused. In this training sample, if the applicant s score is higher than 0, he belongs to the good class, and if lower than 0, belongs to the bad class, see as the table 3:
6 Score Table 3 Training results in two-class 0 <0 Total Class 1( g ood credit) 2( bad credit) Good credit Bad credit Total The bad rate Classification error (type I) (type II) Average error The average error is 34.9%, the type I error, which grant credit to a bad risk applicant, is 28.16%, and type II applicants comprehensive score. By the result of SPSS, we can compute the coefficient of every index, table 4 shows error, which deny credit to a good risk applicant, is 37.09%. the result (for the convenience and corresponding with the comprehensive score, each coefficient all extended 100 (2) The testing results times). As to the prediction to new samples, we should get a predicting function, through which we can compute the new Table 4 the coefficient of each index Index X1 X2 X3 X4 X5 X6 X7 X8 X9 X10 X11 X12 X13 X14 Coeffi cient T hen, we get the comprehensive function as flows: y = 9.49X X X X X X X X X X X X X +7.93X (4) After the data pretreatment and data standardization classify the scores with the same classifying method that has the same treatment as the training sample, we use introduced in train results part. Multiple-class results see the the equation 4 to compute each applicant s score, and table 5 and two-class results see table 6: Table 5 testing results in multiple-class Score 68 34~68 0~34-34~0-68~-34 <-68 Total Class Good credit Bad credit Total The bad rate Odds of good/bad * The score in every class includes the lower value and not includes the upper value
7 Table 6 testing results in two-class Score >0 <0 Total Class 1(good credit) 2(bad credit) Good credit Bad credit Total The bad rate The o dds of good/bad Classif ication error (type I) (typ e II) Average error With the result compared with that of training samples, make a contrast research upon the current scoring method the testing result is better not only in bad person s ratio in the former two classes in multiple-class, but also in the average error in two-class. This shows that the model has a preferable application foreground. How to select appropriate classifying value to get different classes will be based on the stratagems and the business demand of the bank. This can be represented by the credit criteria that bank have selected, strict or lenient. Bank will consider the default rate and the classification error synthetically, that is, based on the risk, set different credit grades, and make out the corresponding credit amount. The comprehensive score presented in this paper will realize the compounding classes under the different default rate and misclassifying rate conveniently. 4. A Comparison Study that one big commercial bank is using in China. The first class means the good credit and will grant the credit to the applicants, and the other class means the bad credit and will refuse the application. The clique value of the bank is 110, which is the lowest score that can grant the credit to the applicants. We get 199 records that belong to the first class and 151 records to the other in the 350 samples by the 110-clique value. To make a reasonable contrast, we classify two classes which have the same records in each class, that is, 199 records in class 1 meaning the good credit class, the other 151 means the bad credit and will refuse their applications. The class comprehensive score is 0.06 now, bigger than 0.06 belongs to the class 1, and the others to the class 2, table 7 shows the detail contrast. We use the test samples which include 350 records to cost of granting credit to a high-risk applicant is Table 7 results contrast with two methods significantly greater than that of denying credit to a good >=110 <110 Total >0.06 risk <0.06 applicant. Therefore, the bank will concentrate more on type I error, and requires a small error rate. The type I 1(good cr edit) 2(bad credit) 1(good credit) 2(bad credit) error of our method is 27.27%, reducing about 40% to dit the bank s result which is 44.32%. If we take a higher it clique value, the type I error will decrease sharply as we can 199 see from the multiple-class 151 classification, which is error (typ e I) (type II) (type outlined I) in (type table 3. This I) contrast of results shows that our ror method has a better effect in practice of credit card risk management. We have the average error 31.71%, much lower than the bank s 40.29%. It is generally believed that the
8 5. Summary and conclusions We present an improved method which is applicable to Chinese commercial banks. It has the following characteristics: (1) Index coefficients are given only based on sample data, thus the objectivity is guaranteed. (2) It is able to adjust the scoring function dynamically if needed, hence has a good adaptability to the population drift phenomenon, etc.. (3) The default rate is increasing monotonously with the decreasing of the score. The banks can set different grades according to the strategy and business with the relationship of score and default rate in order to control the credit card risk. meets the requirement of credit card risk management, and the The primary results of our method manifest that it test result indicated that it has a good application foreground. The demonstration test shows that it is superior to the current method used in Chinese commercial banks. We hope to get better results through more logical data pretreatment and more suitable sample scale selecting etc. Credit scoring and its applications, Society for Industrial and Applied Mathematics, Philadelphia, [3] He Xiaoqun, Modern statistical methods and applications, Ren Ming University of China Publishing Company, (in Chinese) [4] Liu Xianyong, SPSS The statistical soft and its application. National Defence Industry Press, (in Chinese) [5] Zhang Wei, Li yushuang. The overview of the credit risk in commercial bank, the Jounal of Management Science, (in Chinese) [7] Ba Shusong. The risk management function and effects of credit scoring system to commercial band, Congqing Finance, (in Chinese) [8] A Chinese commercial bank. Personal credit scoring criterion and personal credit grade. (in Chinese) [9] Zhang Aiming, etc.. The PCA based prediction model and demonstration research in public companies financial failure. Journal of financial research, (in Chinese) 6. Acknowledgements This research is partly supported by the President Fund of Chinese Academy of Sciences (CAS). The authors are gradeful to Profs. Ji Lei, Chi Hong, Chen Jianming and other colleagues in the Working Group of Financial and Management Science, of the Institute of policy and Management, CAS, for their helpful advices in writing this paper. References [1] Thomas, Lyn, A survey of credit and behavioral scoring: forecasting financial risk of lending to customers, International Journal of Forecasting, Vol. 16, p , [2] Thomas.Lyn, David.Eldelman and Jonathan.Crook,
Empirical Analysis on the impact of working capital management. Hong ming, Chen & Sha Liu
Empirical Analysis on the impact of working capital management Hong ming, Chen & Sha Liu (School Of Economics and Management in Changsha University of Science and Technology Changsha, Hunan, China 410076)
More informationT-test & factor analysis
Parametric tests T-test & factor analysis Better than non parametric tests Stringent assumptions More strings attached Assumes population distribution of sample is normal Major problem Alternatives Continue
More informationTHE USING FACTOR ANALYSIS METHOD IN PREDICTION OF BUSINESS FAILURE
THE USING FACTOR ANALYSIS METHOD IN PREDICTION OF BUSINESS FAILURE Mary Violeta Petrescu Ph. D University of Craiova Faculty of Economics and Business Administration Craiova, Romania Abstract: : After
More informationU.P.B. Sci. Bull., Series C, Vol. 77, Iss. 1, 2015 ISSN 2286 3540
U.P.B. Sci. Bull., Series C, Vol. 77, Iss. 1, 2015 ISSN 2286 3540 ENTERPRISE FINANCIAL DISTRESS PREDICTION BASED ON BACKWARD PROPAGATION NEURAL NETWORK: AN EMPIRICAL STUDY ON THE CHINESE LISTED EQUIPMENT
More informationFactor Analysis. Principal components factor analysis. Use of extracted factors in multivariate dependency models
Factor Analysis Principal components factor analysis Use of extracted factors in multivariate dependency models 2 KEY CONCEPTS ***** Factor Analysis Interdependency technique Assumptions of factor analysis
More informationAn Empirical Study on the Influence of Perceived Credibility of Online Consumer Reviews
An Empirical Study on the Influence of Perceived Credibility of Online Consumer Reviews GUO Guoqing 1, CHEN Kai 2, HE Fei 3 1. School of Business, Renmin University of China, 100872 2. School of Economics
More informationCredit Card Business in Macao and Its Credit Risk Management
Credit Card Business in Macao and Its Credit Risk Management P.S. Un Research and Statistics Department, Monetary Authority of Macao Abstract Credit card holdings and loans have experienced rapid growth
More informationCredit Risk Models. August 24 26, 2010
Credit Risk Models August 24 26, 2010 AGENDA 1 st Case Study : Credit Rating Model Borrowers and Factoring (Accounts Receivable Financing) pages 3 10 2 nd Case Study : Credit Scoring Model Automobile Leasing
More informationAnalysis of Fire Statistics of China: Fire Frequency and Fatalities in Fires
Analysis of Fire Statistics of China: Fire Frequency and Fatalities in Fires FULIANG WANG, SHOUXIANG LU, and CHANGHAI LI State Key Laboratory of Fire Science University of Science and Technology of China
More informationTAM Analysis of College Students Online Banking Brand Selection Factors
ISSN(Print): 2377-0082 ISSN(Online): 2377-0163 EQUILIBRIUM, CHAOS, AND CONERGENCE IN DYNAMICAL NETWORK In Press TAM Analysis of College Students Online Banking Brand Selection Factors Jing Xu*, Xue Liu,
More informationAn empirical study of factor analysis on M & A performance of listed companies of Chinese pharmaceutical industry
Available online www.jocpr.com Journal of Chemical and Pharmaceutical Research, 2014, 6(4):963-968 Research Article ISSN : 0975-7384 CODEN(USA) : JCPRC5 An empirical study of factor analysis on M & A performance
More informationStudy of Lightning Damage Risk Assessment Method for Power Grid
Energy and Power Engineering, 2013, 5, 1478-1483 doi:10.4236/epe.2013.54b280 Published Online July 2013 (http://www.scirp.org/journal/epe) Study of Lightning Damage Risk Assessment Method for Power Grid
More informationAn Application of the Cox Proportional Hazards Model to the Construction of Objective Vintages for Credit in Financial Institutions, Using PROC PHREG
Paper 3140-2015 An Application of the Cox Proportional Hazards Model to the Construction of Objective Vintages for Credit in Financial Institutions, Using PROC PHREG Iván Darío Atehortua Rojas, Banco Colpatria
More informationThe Study of the Impact of E-commerce Participator on Online Reputation
Send Orders for Reprints to reprints@benthamscience.ae 340 The Open Cybernetics & Systemics Journal, 2014, 8, 340-348 Open Access The Study of the Impact of E-commerce Participator on Online Reputation
More informationThe Data Analysis of Primary and Secondary School Teachers Educational Technology Training Result in Baotou City of Inner Mongolia Autonomous Region
, pp.77-84 http://dx.doi.org/10.14257/ijunesst.2015.8.8.08 The Data Analysis of Primary and Secondary School Teachers Educational Technology Training Result in Baotou City of Inner Mongolia Autonomous
More informationOnline Marketing Strategy for Agricultural Supply Chain and Regional Economic Growth based on E-commerce Perspective
, pp.323-332 http://dx.doi.org/10.14257/ijsia.2015.9.10.29 Online Marketing Strategy for Agricultural Supply Chain and Regional Economic Growth based on E-commerce Perspective Hui Yang and Yajuan Zhang*
More informationResearch on the Custom Pricing Model of Online Retail Clothing
www.sciedu.ca/jbar Journal of Business Administration Research Vol. 3, No. 2; 204 Research on the Custom Pricing Model of Online Retail Clothing Hai Liu, Shuaitong Liang & Shouzhong Hu College of fashion,
More informationAnalysis of China Motor Vehicle Insurance Business Trends
Analysis of China Motor Vehicle Insurance Business Trends 1 Xiaohui WU, 2 Zheng Zhang, 3 Lei Liu, 4 Lanlan Zhang 1, First Autho University of International Business and Economic, Beijing, wuxiaohui@iachina.cn
More informationClassification of Chinese Higher Education Institutions
Classification of Chinese Higher Education Institutions December 5, 2006 Professor Nian Cai Liu Institute of Higher Education and Center for World-Class Universities Shanghai Jiao Tong University 1 Outline
More informationA Quantitative Analysis of Chinese Electronic Information Manufacturers International Marketing Performances Under the Influence of R&D Investment
A Quantitative Analysis of Chinese Electronic Information Manufacturers International Marketing Performances Under the Influence of R&D Investment DU Yuping Research Centre of International Trade and Economics,
More informationLin Cui, Yan Chen and Yang Li Art institute, Hebei Normal University of Science and Technology, Qinhuangdao, 066004, China Email: 2067377625@qq.
V ol., No. (06), pp.363-374 http://dx.doi.org/0.457/ijmue.06...35 The Research on the Influence of Computer Music Perceived Value on Music Perception for Colleges under Different Backgrounds Based on SPSS
More informationNonparametric adaptive age replacement with a one-cycle criterion
Nonparametric adaptive age replacement with a one-cycle criterion P. Coolen-Schrijner, F.P.A. Coolen Department of Mathematical Sciences University of Durham, Durham, DH1 3LE, UK e-mail: Pauline.Schrijner@durham.ac.uk
More informationCONCEPTUAL MODEL OF MULTI-AGENT BUSINESS COLLABORATION BASED ON CLOUD WORKFLOW
CONCEPTUAL MODEL OF MULTI-AGENT BUSINESS COLLABORATION BASED ON CLOUD WORKFLOW 1 XINQIN GAO, 2 MINGSHUN YANG, 3 YONG LIU, 4 XIAOLI HOU School of Mechanical and Precision Instrument Engineering, Xi'an University
More informationStudy of Corporate Governance on Relation between Self-interest
Study of Corporate Governance on Relation between Self-interest Incentive and Cost Stickiness 1 1, First Author College of Accounting, Shanxi University of Finance and Economics, China, E-mail:13834551860@139.com
More informationCredit Risk Analysis Using Logistic Regression Modeling
Credit Risk Analysis Using Logistic Regression Modeling Introduction A loan officer at a bank wants to be able to identify characteristics that are indicative of people who are likely to default on loans,
More informationResearch on the UHF RFID Channel Coding Technology based on Simulink
Vol. 6, No. 7, 015 Research on the UHF RFID Channel Coding Technology based on Simulink Changzhi Wang Shanghai 0160, China Zhicai Shi* Shanghai 0160, China Dai Jian Shanghai 0160, China Li Meng Shanghai
More informationFACTORS AFFECTING CUSTOMERS BUYING DECISIONS OF MOBILE PHONE: A STUDY ON KHULNA CITY, BANGLADESH
FACTORS AFFECTING CUSTOMERS BUYING DECISIONS OF MOBILE PHONE: A STUDY ON KHULNA CITY, BANGLADESH Md Reaz Uddin 1 Nusrat Zahan Lopa 2 and Md. Oheduzzaman 3 1 Assistant Professor, Business Administration
More informationStudents' Opinion about Universities: The Faculty of Economics and Political Science (Case Study)
Cairo University Faculty of Economics and Political Science Statistics Department English Section Students' Opinion about Universities: The Faculty of Economics and Political Science (Case Study) Prepared
More informationEmpirical Research on Influencing Factors of Human Resources Management Outsourcing Degree *
ibusiness, 2013, 5, 168-174 http://dx.doi.org/10.4236/ib.2013.53b036 Published Online September 2013 (http://www.scirp.org/journal/ib) Empirical Research on Influencing Factors of Human Resources Management
More informationDiscussion on Information Asymmetry in B2C E-commerce
Discussion on Information Asymmetry in B2C E-commerce QIN Dezhi 1, ZOU Lifang 2 1. School of Business & Tourism Management, Yunnan University, P.H.D, No.2 North Cuihu Road Kunming, 650091, 86-13888673168
More informationBehavior Model to Capture Bank Charge-off Risk for Next Periods Working Paper
1 Behavior Model to Capture Bank Charge-off Risk for Next Periods Working Paper Spring 2007 Juan R. Castro * School of Business LeTourneau University 2100 Mobberly Ave. Longview, Texas 75607 Keywords:
More informationRandom forest algorithm in big data environment
Random forest algorithm in big data environment Yingchun Liu * School of Economics and Management, Beihang University, Beijing 100191, China Received 1 September 2014, www.cmnt.lv Abstract Random forest
More informationFACTOR ANALYSIS. Factor Analysis is similar to PCA in that it is a technique for studying the interrelationships among variables.
FACTOR ANALYSIS Introduction Factor Analysis is similar to PCA in that it is a technique for studying the interrelationships among variables Both methods differ from regression in that they don t have
More informationAn Analysis of Agricultural Risk and Intelligent Monitoring Technology Fantao Kong 1, a, Shiwei Xu 2,b, Shengwei Wang 3,c and Haipeng Yu 4,d
Advanced Materials Research Vol. 628 (2013) pp 265-269 Online available since 2012/Dec/27 at www.scientific.net (2013) Trans Tech Publications, Switzerland doi:10.4028/www.scientific.net/amr.628.265 An
More informationStudy on the Evaluation for the Knowledge Sharing Efficiency of the Knowledge Service Network System in Agile Supply Chain
Send Orders for Reprints to reprints@benthamscience.ae 384 The Open Cybernetics & Systemics Journal, 2015, 9, 384-389 Open Access Study on the Evaluation for the Knowledge Sharing Efficiency of the Knowledge
More informationTraffic Behavior Analysis with Poisson Sampling on High-speed Network 1
Traffic Behavior Analysis with Poisson Sampling on High-speed etwork Guang Cheng Jian Gong (Computer Department of Southeast University anjing 0096, P.R.China) Abstract: With the subsequent increasing
More informationResearch of Female Consumer Behavior in Cosmetics Market Case Study of Female Consumers in Hsinchu Area Taiwan
usiness, 2010, 2, 348-353 doi:10.4236/ib.2010.24045 Published Online December 2010 (http://www.scirp.org/journal/ib) Research of Female Consumer Behavior in Cosmetics Market Case Study of Female Consumers
More informationImproving the Performance of Data Mining Models with Data Preparation Using SAS Enterprise Miner Ricardo Galante, SAS Institute Brasil, São Paulo, SP
Improving the Performance of Data Mining Models with Data Preparation Using SAS Enterprise Miner Ricardo Galante, SAS Institute Brasil, São Paulo, SP ABSTRACT In data mining modelling, data preparation
More informationAn Empirical Study of Influential Factors of Debt Financing
ISSN 1479-3889 (print), 1479-3897 (online) International Journal of Nonlinear Science Vol.3(2007) No.3,pp.208-212 An Empirical Study of Influential Factors of Debt Financing Jing Wu School of Management,
More informationBinary Logistic Regression
Binary Logistic Regression Main Effects Model Logistic regression will accept quantitative, binary or categorical predictors and will code the latter two in various ways. Here s a simple model including
More information5.2 Customers Types for Grocery Shopping Scenario
------------------------------------------------------------------------------------------------------- CHAPTER 5: RESULTS AND ANALYSIS -------------------------------------------------------------------------------------------------------
More informationFactor Analysis. Chapter 420. Introduction
Chapter 420 Introduction (FA) is an exploratory technique applied to a set of observed variables that seeks to find underlying factors (subsets of variables) from which the observed variables were generated.
More informationGeneralized Linear Models
Generalized Linear Models We have previously worked with regression models where the response variable is quantitative and normally distributed. Now we turn our attention to two types of models where the
More informationCREDIT RISK ASSESSMENT FOR MORTGAGE LENDING
IMPACT: International Journal of Research in Business Management (IMPACT: IJRBM) ISSN(E): 2321-886X; ISSN(P): 2347-4572 Vol. 3, Issue 4, Apr 2015, 13-18 Impact Journals CREDIT RISK ASSESSMENT FOR MORTGAGE
More informationAccurately and Efficiently Measuring Individual Account Credit Risk On Existing Portfolios
Accurately and Efficiently Measuring Individual Account Credit Risk On Existing Portfolios By: Michael Banasiak & By: Daniel Tantum, Ph.D. What Are Statistical Based Behavior Scoring Models And How Are
More informationJournal of Chemical and Pharmaceutical Research, 2014, 6(3):34-39. Research Article. Analysis of results of CET 4 & CET 6 Based on AHP
Available online www.jocpr.com Journal of Chemical and Pharmaceutical Research, 2014, 6(3):34-39 Research Article ISSN : 0975-7384 CODEN(USA) : JCPRC5 Analysis of results of CET 4 & CET 6 Based on AHP
More informationThe Application of 360 +KPI Performance Evaluation Model in Chinese and Western Culture Background
The Application of 360 +KPI Performance Evaluation Model in Chinese and Western Culture Background Wei Tan The Institute of Public Affairs, Chongqing Three Gorges University The No.780 of ShaLong street,
More informationChina s Social Medical Insurance System Needs the Support of a Sound Medical Service System
Network of Asia-Pacific Schools and Institutes of Public Administration and Governance (NAPSIPAG) Annual Conference 2005 BEIJING, PRC, 5-7 DECEMBER 2005 THEME: THE ROLE OF PUBLIC ADMINISTRATION IN BUILDING
More informationDoes organizational culture cheer organizational profitability? A case study on a Bangalore based Software Company
Does organizational culture cheer organizational profitability? A case study on a Bangalore based Software Company S Deepalakshmi Assistant Professor Department of Commerce School of Business, Alliance
More informationEvaluation Model of Buyers Dynamic Reputation in E-commerce
Vol.0, No.5 (05), pp.53-64 http://dx.doi.org/0.457/ijmue.05.0.5.07 Evaluation Model of Buyers Dynamic Reputation in E-commerce Yang Zhang ), Tieying Liu ), Ru Li 3) and Zhenhai Wan 4) School of Computer
More informationResearch on Trust Management Strategies in Cloud Computing Environment
Journal of Computational Information Systems 8: 4 (2012) 1757 1763 Available at http://www.jofcis.com Research on Trust Management Strategies in Cloud Computing Environment Wenjuan LI 1,2,, Lingdi PING
More informationComparison of sales forecasting models for an innovative agro-industrial product: Bass model versus logistic function
The Empirical Econometrics and Quantitative Economics Letters ISSN 2286 7147 EEQEL all rights reserved Volume 1, Number 4 (December 2012), pp. 89 106. Comparison of sales forecasting models for an innovative
More informationBasic Concepts in Research and Data Analysis
Basic Concepts in Research and Data Analysis Introduction: A Common Language for Researchers...2 Steps to Follow When Conducting Research...3 The Research Question... 3 The Hypothesis... 4 Defining the
More informationCUSTOMER PERCEPTION AND RESPONSE TOWARDS ONLINE MARKETING IN CHENNAI CITY
CUSTOMER PERCEPTION AND RESPONSE TOWARDS ONLINE MARKETING IN CHENNAI CITY Dr. K. KRISHNAMURTHY Assistant Professor, Research Supervisor, P.G & Research Dept of Commerce, Rajeswari Vedachalam Government
More informationThe Analysis on Employability Gap of Students in China Vocational-Technical School and Its Causes
The Analysis on Employability Gap of Students in China Vocational-Technical School and Its Causes WANG Ting Business School of China University of Political Science and Law, Beijing, P.R.China, 102249
More informationEmpirical Analysis on Urban Retail Business Spatial Distribution Influence Factors
Empirical Analysis on Urban Retail Business Spatial Distribution Influence Factors ZHOU Chunhua, CHEN Jiandong School of Economics, University of Jinan, Shandong, 250022 Abstract: The development of urban
More informationResearch on Incubation Performance of China Information Technology Business Incubators with DEA
Journal of Modeling and Optimization 7:1 (2015) Research on Incubation Performance of China Information Technology Business Incubators with DEA Jiangping Wan 1, Guangwei Pan 2,Lianyu Liang 3 1. School
More informationNCSS Statistical Software Principal Components Regression. In ordinary least squares, the regression coefficients are estimated using the formula ( )
Chapter 340 Principal Components Regression Introduction is a technique for analyzing multiple regression data that suffer from multicollinearity. When multicollinearity occurs, least squares estimates
More informationExploratory Factor Analysis of Demographic Characteristics of Antenatal Clinic Attendees and their Association with HIV Risk
Doi:10.5901/mjss.2014.v5n20p303 Abstract Exploratory Factor Analysis of Demographic Characteristics of Antenatal Clinic Attendees and their Association with HIV Risk Wilbert Sibanda Philip D. Pretorius
More informationSuccession planning in Chinese family-owned businesses in Hong Kong: an exploratory study on critical success factors and successor selection criteria
Succession planning in Chinese family-owned businesses in Hong Kong: an exploratory study on critical success factors and successor selection criteria By Ling Ming Chan BEng (University of Newcastle upon
More informationEvaluating Effect of Free Float on Liquidity Increase, Depth and Efficiency of Companies Listed in Tehran Stock Exchange
2014 (4): 1-7 Evaluating Effect of Free Float on Liquidity Increase, Depth and Efficiency of Companies Listed in Tehran Stock Exchange Mohammad Reza Shahrasbi 1, Dr. Mehdi Taghavi 2, Dr. Kambiz Hozhabr
More informationEconomic Analysis on Development of Marine Insurance in Shanghai. 1 Introduction
Economic Analysis on Development of Marine Insurance in Shanghai WANG Yun, LIU Juanjuan Research Institute for Science of Water Transport Economy, Shanghai Maritime University, 155 Pudong Av., Shanghai,
More informationSimple Linear Regression Inference
Simple Linear Regression Inference 1 Inference requirements The Normality assumption of the stochastic term e is needed for inference even if it is not a OLS requirement. Therefore we have: Interpretation
More informationAn Empirical Examination of the Relationship between Financial Trader s Decision-making and Financial Software Applications
Empirical Examination of Financial decision-making & software apps. An Empirical Examination of the Relationship between Financial Trader s Decision-making and Financial Software Applications Research-in-Progress
More informationCopula model estimation and test of inventory portfolio pledge rate
International Journal of Business and Economics Research 2014; 3(4): 150-154 Published online August 10, 2014 (http://www.sciencepublishinggroup.com/j/ijber) doi: 10.11648/j.ijber.20140304.12 ISS: 2328-7543
More informationWhen to Refinance Mortgage Loans in a Stochastic Interest Rate Environment
When to Refinance Mortgage Loans in a Stochastic Interest Rate Environment Siwei Gan, Jin Zheng, Xiaoxia Feng, and Dejun Xie Abstract Refinancing refers to the replacement of an existing debt obligation
More informationChinese Local Government Performance Appraisal Based on KPI
2011 International Conference on Information Management and Engineering (ICIME 2011) IPCSIT vol. 52 (2012) (2012) IACSIT Press, Singapore DOI: 10.7763/IPCSIT.2012.V52.3 Chinese Local Government Performance
More informationResearch on small and medium enterprises financing mode based on supply chain finance
Available online www.jocpr.com Journal of Chemical and Pharmaceutical Research, 0, ():88-8 Research Article ISSN : 09-8 CODEN(USA) : JCPRC Research on small and medium enterprises financing mode based
More informationOverview of Factor Analysis
Overview of Factor Analysis Jamie DeCoster Department of Psychology University of Alabama 348 Gordon Palmer Hall Box 870348 Tuscaloosa, AL 35487-0348 Phone: (205) 348-4431 Fax: (205) 348-8648 August 1,
More informationCredit risk management of commercial bank
Available online www.jocpr.com Journal of Chemical and Pharmaceutical Research, 2014, 6(5):1784-1788 Research Article ISSN : 0975-7384 CODEN(USA) : JCPRC5 Credit risk management of commercial bank Qian
More informationFactor Analysis. Advanced Financial Accounting II Åbo Akademi School of Business
Factor Analysis Advanced Financial Accounting II Åbo Akademi School of Business Factor analysis A statistical method used to describe variability among observed variables in terms of fewer unobserved variables
More informationBig Data Analytics of Multi-Relationship Online Social Network Based on Multi-Subnet Composited Complex Network
, pp.273-284 http://dx.doi.org/10.14257/ijdta.2015.8.5.24 Big Data Analytics of Multi-Relationship Online Social Network Based on Multi-Subnet Composited Complex Network Gengxin Sun 1, Sheng Bin 2 and
More informationIMPLICATIONS OF LOGISTIC SERVICE QUALITY ON THE SATISFACTION LEVEL AND RETENTION RATE OF AN E-COMMERCE RETAILER S CUSTOMERS
Associate Professor, Adrian MICU, PhD Dunarea de Jos University of Galati E-mail: mkdradrianmicu@yahoo.com Professor Kamer AIVAZ, PhD Ovidius University of Constanta Alexandru CAPATINA, PhD Dunarea de
More informationReal estate investment project risk analysis
Available online www.jocpr.com Journal of Chemical and Pharmaceutical Research, 2014, 6(5):1789-1794 Research Article ISSN : 0975-7384 CODEN(USA) : JCPRC5 Real estate investment project risk analysis Li
More informationPREDICTION FINANCIAL DISTRESS BY USE OF LOGISTIC IN FIRMS ACCEPTED IN TEHRAN STOCK EXCHANGE
PREDICTION FINANCIAL DISTRESS BY USE OF LOGISTIC IN FIRMS ACCEPTED IN TEHRAN STOCK EXCHANGE * Havva Baradaran Attar Moghadas 1 and Elham Salami 2 1 Lecture of Accounting Department of Mashad PNU University,
More informationEmpirical Research on the Influence Factors of E-commerce Development in China
Send Orders for Reprints to reprints@benthamscience.ae 76 The Open Cybernetics & Systemics Journal, 2015, 9, 76-82 Open Access Empirical Research on the Influence Factors of E-commerce Development in China
More informationAN ANALYSIS OF THE FINANCIAL CHARACTERISTICS OF FIRMS THAT EXPERIENCED HEAVY BUYING AND SELLING BY INSIDERS DURING A PERIOD OF ECONOMIC RECESSION
An Analysis of the Financial Characteristics of Firms that Experienced Heavy Buying and Selling by Insiders During a Period of Economic Recession AN ANALYSIS OF THE FINANCIAL CHARACTERISTICS OF FIRMS THAT
More informationFactors affecting teaching and learning of computer disciplines at. Rajamangala University of Technology
December 2010, Volume 7, No.12 (Serial No.73) US-China Education Review, ISSN 1548-6613, USA Factors affecting teaching and learning of computer disciplines at Rajamangala University of Technology Rungaroon
More informationStudy on Human Performance Reliability in Green Construction Engineering
Study on Human Performance Reliability in Green Construction Engineering Xiaoping Bai a, Cheng Qian b School of management, Xi an University of Architecture and Technology, Xi an 710055, China a xxpp8899@126.com,
More informationIntroduction to consumer credit and credit scoring
University Press Scholarship Online You are looking at 1-10 of 12 items for: keywords : credit scoring Credit scoring Tony Van Gestel and Bart Baesens in Credit Risk Management: Basic Concepts: Financial
More informationMethod of Fault Detection in Cloud Computing Systems
, pp.205-212 http://dx.doi.org/10.14257/ijgdc.2014.7.3.21 Method of Fault Detection in Cloud Computing Systems Ying Jiang, Jie Huang, Jiaman Ding and Yingli Liu Yunnan Key Lab of Computer Technology Application,
More informationTHE USE OF PERFORMANCE MEASURES IN SMALL TO MEDIUM ENTERPRISES (SMEs) AN EXPLORATORY STUDY
THE USE OF PERFORMANCE MEASURES IN SMALL TO MEDIUM ENTERPRISES (SMEs) AN EXPLORATORY STUDY Brendan Phillips, Thomas Tan Tsu Wee and Tekle Shanka Curtin Business School Keywords: SMEs, performance measures,
More informationStatistical properties of trading activity in Chinese Stock Market
Physics Procedia 3 (2010) 1699 1706 Physics Procedia 00 (2010) 1 8 Physics Procedia www.elsevier.com/locate/procedia Statistical properties of trading activity in Chinese Stock Market Xiaoqian Sun a, Xueqi
More informationDemand Forecasting Optimization in Supply Chain
2011 International Conference on Information Management and Engineering (ICIME 2011) IPCSIT vol. 52 (2012) (2012) IACSIT Press, Singapore DOI: 10.7763/IPCSIT.2012.V52.12 Demand Forecasting Optimization
More informationCHAPTER 5: CONSUMERS ATTITUDE TOWARDS ONLINE MARKETING OF INDIAN RAILWAYS
CHAPTER 5: CONSUMERS ATTITUDE TOWARDS ONLINE MARKETING OF INDIAN RAILWAYS 5.1 Introduction This chapter presents the findings of research objectives dealing, with consumers attitude towards online marketing
More informationBusiness Performance Evaluation Model for the Taiwan Electronic Industry based on Factor Analysis and AHP Method
Proceedings of the World Congress on Engineering 200 Vol III WCE 200, June 0 - July 2, 200, London, U.K. Business Performance Evaluation Model for the Taiwan Electronic Industry based on Factor Analysis
More informationFACTOR ANALYSIS NASC
FACTOR ANALYSIS NASC Factor Analysis A data reduction technique designed to represent a wide range of attributes on a smaller number of dimensions. Aim is to identify groups of variables which are relatively
More informationLearning Attitude and Its Effect on Applying Cloud Computing Service to IT Education
Learning Attitude and Its Effect on Applying Cloud Computing Service to IT Education Chen-Feng Wu Department of Information Management, Yu Da University No 168, Hsueh-fu Rd, Tanwen Village, Chaochiao Township,
More informationE-R Method Applied to Design the Teacher Information Management System s Database Model
Vol. 6, o. 4, August, 2013 E-R ethod Applied to Design the Teacher Information anagement System s Database odel Yingjian Kang 1 and Dan Zhao 2 1 Department of Computer Technology, College of Telecommunications
More informationLIST OF TABLES. 4.3 The frequency distribution of employee s opinion about training functions emphasizes the development of managerial competencies
LIST OF TABLES Table No. Title Page No. 3.1. Scoring pattern of organizational climate scale 60 3.2. Dimension wise distribution of items of HR practices scale 61 3.3. Reliability analysis of HR practices
More informationTHE APPLICATION OF DATA MINING TECHNOLOGY IN REAL ESTATE MARKET PREDICTION
THE APPLICATION OF DATA MINING TECHNOLOGY IN REAL ESTATE MARKET PREDICTION Xian Guang LI, Qi Ming LI Department of Construction and Real Estate, South East Univ,,Nanjing, China. Abstract: This paper introduces
More informationMULTIPLE REGRESSION ANALYSIS OF MAIN ECONOMIC INDICATORS IN TOURISM. R, analysis of variance, Student test, multivariate analysis
Journal of tourism [No. 8] MULTIPLE REGRESSION ANALYSIS OF MAIN ECONOMIC INDICATORS IN TOURISM Assistant Ph.D. Erika KULCSÁR Babeş Bolyai University of Cluj Napoca, Romania Abstract This paper analysis
More informationharpreet@utdallas.edu, {ram.gopal, xinxin.li}@business.uconn.edu
Risk and Return of Investments in Online Peer-to-Peer Lending (Extended Abstract) Harpreet Singh a, Ram Gopal b, Xinxin Li b a School of Management, University of Texas at Dallas, Richardson, Texas 75083-0688
More informationGlossary of Terms Ability Accommodation Adjusted validity/reliability coefficient Alternate forms Analysis of work Assessment Battery Bias
Glossary of Terms Ability A defined domain of cognitive, perceptual, psychomotor, or physical functioning. Accommodation A change in the content, format, and/or administration of a selection procedure
More informationPolynomial Operations and Factoring
Algebra 1, Quarter 4, Unit 4.1 Polynomial Operations and Factoring Overview Number of instructional days: 15 (1 day = 45 60 minutes) Content to be learned Identify terms, coefficients, and degree of polynomials.
More informationAnalysis of Electricity Price Policy and Economic Growth
Journal of Scientific & Industrial Research Vol 74, January 2015, pp. 11-18 Analysis of Electricity Price Policy and Economic Growth Weida He 1, Chuan Zhang 1 and Rong Hao 1,2 * 1 Donlinks School of Economics
More informationAN ANALYSIS OF FOOD SAFETY MANAGEMENT SYSTEMS CERTIFICATION: THE PORTUGUESE CASE
AN ANALYSIS OF FOOD SAFETY MANAGEMENT SYSTEMS CERTIFICATION: THE PORTUGUESE CASE Sofia Teixeira, MSc Student, University of Porto, sofiatteixeira@gmail.com Paulo Sampaio, University of Minho, paulosampaio@dps.uminho.pt
More informationInternational Journal of Business and Social Science Vol. 3 No. 3; February 2012
50 COMOVEMENTS AND STOCK MARKET INTEGRATION BETWEEN INDIA AND ITS TOP TRADING PARTNERS: A MULTIVARIATE ANALYSIS OF INTERNATIONAL PORTFOLIO DIVERSIFICATION Abstract Alan Harper, Ph.D. College of Business
More informationCommon factor analysis
Common factor analysis This is what people generally mean when they say "factor analysis" This family of techniques uses an estimate of common variance among the original variables to generate the factor
More informationTHE UTILISATION OF STATISTICAL METHODS IN THE AREA OF INVENTORY MANAGEMENT. Petr BESTA, Kamila JANOVSKÁ, Šárka VILAMOVÁ, Iveta VOZŇÁKOVÁ, Josef KUTÁČ
THE UTILISATION OF STATISTICAL METHODS IN THE AREA OF INVENTORY MANAGEMENT Petr BESTA, Kamila JANOVSKÁ, Šárka VILAMOVÁ, Iveta VOZŇÁKOVÁ, Josef KUTÁČ VSB Technical University of Ostrava, Ostrava, Czech
More information