Spelling suggestions: "subject:"discriminant analysis"" "subject:"oiscriminant analysis""
11 |
Feature extraction VIA kernel weighted discriminant analysis methods /Dai, Guang. January 2007 (has links)
Thesis (M.Phil.)--Hong Kong University of Science and Technology, 2007. / Includes bibliographical references (leaves 83-90). Also available in electronic version.
|
12 |
Principles and methodology of non-parametric discrimination /Wong, Tat-yan. January 1981 (has links)
Thesis--M. Phil., University of Hong Kong, 1982.
|
13 |
An analysis of repeated measurements on experimental units in a two-way classification /McNee, Richard Cameron, January 1966 (has links)
Thesis (M.S.)--Virginia Polytechnic Institute, 1966. / Vita. Abstract. Includes bibliographical references (leaf 51). Also available via the Internet.
|
14 |
Principles and methodology of non-parametric discriminationWong, Tat-yan. January 1981 (has links)
Thesis, M.Phil., University of Hong Kong, 1982. / Also available in print.
|
15 |
Assessing the influence of observations on the generalization performance of the generalization performance of the Kernel Fisher Discriminant Classifier /Lamont, Morné Michael Connell. January 2008 (has links)
Dissertation (PhD)--University of Stellenbosch, 2008. / Bibliography. Also available via the Internet.
|
16 |
Feature selection and discriminant analysis in data mining /Youn, Eun Seog. January 2004 (has links)
Thesis (Ph. D.)--University of Florida, 2004. / Includes vita. Includes bibliographical references. Also available via the World Wide Web (viewed June 30, 2006).
|
17 |
Continuous methods in optimization and its application in discriminant analysisZhang, Leihong 01 January 2008 (has links)
No description available.
|
18 |
Some Results in Classification TheoryHart, Bradd January 1986 (has links)
Note:
|
19 |
Modelling and analysis of ranking data with misclassification.January 2007 (has links)
Chan, Ho Wai. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2007. / Includes bibliographical references (leaves 56). / Abstracts in English and Chinese. / Abstract --- p.ii / Acknowledgement --- p.iv / Chapter 1 --- Introduction --- p.1 / Chapter 2 --- Model --- p.3 / Chapter 3 --- Implementation by Mx --- p.10 / Chapter 3.1 --- Example 1 --- p.10 / Chapter 3.2 --- Example 2 --- p.22 / Chapter 4 --- Covariance structure analysis --- p.26 / Chapter 5 --- Simulation --- p.29 / Chapter 5.1 --- Simulation 1 --- p.29 / Chapter 5.2 --- Simulation 2 --- p.36 / Chapter 6 --- Discussion --- p.41 / Appendix A: Mx input script for ranking data data with p =4 --- p.43 / Appendix B: Selection matrices for ranking data with p = 4 --- p.47 / Appendix C: Mx input script for ranking data data with p = 3 --- p.50 / Appendix D: Mx input script for p = 4 with covariance structure --- p.53 / References --- p.56
|
20 |
Investor Risk Tolerance: Testing The Efficacy Of Demographics As Differentiating and Classifying FactorsGrable, John E. 29 October 1997 (has links)
This study was designed to determine whether the variables gender, age, marital status, occupation, self-employment, income, race, and education could be used individually or in combination to both differentiate among levels of investor risk tolerance and classify individuals into risk-tolerance categories. The Leimberg, Satinsky, LeClair, and Doyle (1993) financial management model was used as the theoretical basis for this study. The model explains the process of how investment managers effectively develop plans to allocate a client's scarce investment resources to meet financial objectives.
An empirical model for categorizing investors into risk-tolerance categories using demographic factors was developed and empirically tested using data from the 1992 Survey of Consumer Finances (SCF) (N = 2,626). The average respondent was affluent and best represented the profile of an investment management client.
Based on findings from a multiple discriminant analysis test it was determined that respondent demographic characteristics were significant in differentiating among levels of risk tolerance at the p < .0001 level (i.e., gender, married, single but previously married, professional occupational status, self-employment status, income, White, Black, and Hispanic racial background, and educational level), while three demographic characteristics were found to be statistically insignificant (i.e., age, Asian racial background, and never married). Multiple discriminant analysis also revealed that the demographic variables examined in this study explained approximately 20% of the variance among the three levels of investor risk tolerance.
Classification equations were generated. The classification procedure offered only a 20% improvement-over-chance, which was determined to be a low proportional reduction in error. The classification procedure also generated unacceptable levels of false positive classifications, which led to over classification of respondents into high and no risk-tolerance categories, while under classifying respondents into the average risk-tolerance category.
Two demographic characteristics were determined to be the most effective in differentiating among and classifying respondents into risk-tolerance categories. Classes of risk tolerance differed most widely on respondents' educational level and gender. Educational level of respondents was determined to be the most significant optimizing factor. It also was concluded that demographic characteristics provide only a starting point in assessing investor risk tolerance. Understanding risk tolerance is a complicated process that goes beyond the exclusive use of demographic characteristics. More research is needed to determine which additional factors can be used by investment managers to increase the explained variance in risk-tolerance differences. / Ph. D.
|
Page generated in 0.0617 seconds