abalone数据介绍.docVIP

1. Title of Database: Abalone data 2. Sources: (a) Original owners of database: Marine Resources Division Marine Research Laboratories - Taroona Department of Primary Industry and Fisheries, Tasmania GPO Box 619F, Hobart, Tasmania 7001, Australia (contact: Warwick Nash +61 02 277277, wnash@dpi.tas.gov.au) (b) Donor of database: Sam Waugh (Sam.Waugh@cs.utas.edu.au) Department of Computer Science, University of Tasmania GPO Box 252C, Hobart, Tasmania 7001, Australia (c) Date received: December 1995 3. Past Usage: Sam Waugh (1995) Extending and benchmarking Cascade-Correlation, PhD thesis, Computer Science Department, University of Tasmania. -- Test set performance (final 1044 examples, first 3133 used for training): 24.86% Cascade-Correlation (no hidden nodes) 26.25% Cascade-Correlation (5 hidden nodes) 21.5% C4.5 0.0% Linear Discriminate Analysis 3.57% k=5 Nearest Neighbour (Problem encoded as a classification task) -- Data set samples are highly overlapped. Further information is required to separate completely using affine combinations. Other restrictions to data set examined. David Clark, Zoltan Schreter, Anthony Adams A Quantitative Comparison of Dystal and Backpropagation, submitted to the Australian Conference on Neural Networks (ACNN96). Data set treated as a 3-category classification problem (grouping ring classes 1-8, 9 and 10, and 11 on). -- Test set performance (3133 training, 1044 testing as above): 64% Backprop 55% Dystal -- Previous work (Waugh, 1995) on same data set: 61.40% Cascade-Correlation (no hidden nodes) 65.61% Cascade-Correlation (5 hidden nodes) 59.2% C4.5 32.57% Linear Discriminate Analysis 62.46% k=5 Nearest Neighbour 4. Relevant Information Paragraph: Predicting the age of abalone from physical measurements. The age of abalone is determined by cutting the shell through the cone, staining it, and counting the number of rings through a microscope -- a b