Evaluation of Regression Splines: A Multi-criteria Decision Analysis Approach

Approaches to analyzing statistical data can be classified as either confirmatory or exploratory. Confirmatory data analysis requires the explicit specification of one or more hypotheses by the researcher followed by the testing of these hypotheses. In this project we aim to develop a knowledge discovery via data mining (KDDM) process model based context-aware multi-criteria framework for selecting the most appropriate causal explanatory model based on the researchers subjective preferences including accuracy, simplicity, the relative importance of variables in his/her tentative research model, relative preferences for inclusion of some causal relationships.