Combining heterogeneous classifiers for stock selection
Albanis, George T. and Batchelor, R. A. (1999) Combining heterogeneous classifiers for stock selection. Working Paper. Coventry: Warwick Business School, Financial Econometrics Research Centre. Working papers (Warwick Business School. Financial Econometrics Research Centre) (No.02-).
WRAP_Albanis_fwp02-01.pdf - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
Official URL: http://www2.warwick.ac.uk/fac/soc/wbs/research/wfr...
Combining unbiased forecasts of continuous variables necessarily reduces the error variance below that of the median individual forecast. However, this does not necessarily hold for forecasts of discrete variables, or where the costs of errors are not directly related to the error variance. This paper investigates empirically the benefits of combining forecasts of outperforming shares, based on five linear and nonlinear statistical classification techniques, including neural network and recursive partitioning methods. We find that simple “Majority Voting” improves accuracy and profitability only marginally. Much greater gains come from applying the “Unanimity Principle”, whereby a share is not held in the high-performing portfolio unless all classifiers agree.
|Item Type:||Working or Discussion Paper (Working Paper)|
|Subjects:||H Social Sciences > HG Finance
H Social Sciences > HB Economic Theory
|Divisions:||Faculty of Social Sciences > Warwick Business School > Financial Econometrics Research Centre
Faculty of Social Sciences > Warwick Business School
|Library of Congress Subject Headings (LCSH):||Economic forecasting, Stock exchange, Discriminant analysis, Recursive partitioning|
|Series Name:||Working papers (Warwick Business School. Financial Econometrics Research Centre)|
|Publisher:||Warwick Business School, Financial Econometrics Research Centre|
|Place of Publication:||Coventry|
|Official Date:||September 1999|
|Number of Pages:||31|
|Status:||Not Peer Reviewed|
|Access rights to Published version:||Open Access|
Albanis G.T., and R. A. Batchelor, 1999a, Using probabilistic neural networks and rule induction
Actions (login required)