NCJ Number
170587
Journal
Substance Use and Misuse Volume: 33 Issue: 3 Dated: (1998) Pages: 709-716
Date Published
1998
Length
8 pages
Annotation
The characteristics, applications, and results of Linear Discrimination Analysis (LDA) and Artificial Neural Networks (ANN) were examined and applied to the classification of 1,342 people in 12 towns in the Lazio region of Italy.
Abstract
The database used 124 variables; the ANN reduced the number of variables to 48. The participants provided data on their demographics, education and professional status, work and study, territorial mobility, the quality of urban life, social relationships and spare time, and indicators of economic status and consumption. The research assigned a weight to each town in proportion to its own immigration in comparison to the total immigration in Lazio during the time period selected. Results revealed that ANNs were superior to LDA in their classification capabilities in this case. The two approaches have both strong and weak aspects. Tables, author biography and photograph, list of software used, and 24 references