Iterative Dichotomiser Decision Tree for Risk Analysis in Innovation Management

2. Research Methodology

The framework is organized in three phases. In preliminary phase, the input data for classification is collected from historical literature. Each innovative activity is characterized by a tuple (x, y) where X is the attribute set and y is a special attribute, designated as the class label or target attribute. In this phase, database is prepared for next phase. In second phase, decision tree is drawn based on target attribute that is “risk class” here using ID3 algorithm. In the last phase, some proper rules are discovered from the tree and are explained for providing appropriate information for predictors, investors and innovators. Following, these triple phases are explained in detail.