GEOSCAN Search Results: Fastlink


TitleA decision-tree approach to mineral potential mapping in Snow Lake area, Manitoba
AuthorReddy, R K T; Bonham-Carter, G F
SourceCanadian Journal of Remote Sensing vol. 17, no. 2, 1991 p. 191-200,
Alt SeriesGeological Survey of Canada, Contribution Series 19290
PublisherInforma UK Limited
Mediapaper; on-line; digital
File formatpdf
AreaSnow Lake
Subjectsmineral potential; mapping techniques
Released2014 08 01
Decision-tree analysis (DTA) has been used to predict zones favourable for base-metal mineralization in the Snow Lake area, Manitoba. A multi-layer database was created using a geographic information system (GIS), with maps of geophysical, geochemical, geological, remotely-sensed and mineral occurrence variables. An attribute file was produced by creating a multi-map overlay where the newly-created polygons were the records or cases, and the maps were the variables. This attribute file was imported into a software package for building decision trees. During the analysis, the dataset was recursively partitioned into mutually exclusive, exhaustive subsets which best describe the response variable. Using the Snow Lake data, the presence/absence of base-metal occurrences in each polygon was treated as the categorical response variable and the predictor variables included magnetic and gravity values, lithological map units and proximity to particular formation contacts. Output from DTA was a tree and an associated set of decision rules. Each end node in the tree corresponds to a rule, defining a class or subset of the polygons. The spatial distribution of these subsets was mapped with the GIS to show areas with varying levels and types of base-metal potential. DTA is a type of exploratory data analysis that avoids some of the assumptions of Bayesian weights of evidence or general linear models.

Date modified: