2. Random Forest - Recall - 76, Accuracy - 73
3. Gradient Boosting Machine- Recall - 70, Accuracy - 74
Patterns extraction from decision trees (C5.O)
- Using the same preprocessed data which has been used above, built a decision tree algorithm using C50 to build a classifier.
From the summary of the above model, rules have been extracted which give the information about the
a)Lift
b)Confidence
c)Support
d) Condition
Analysis
The experimental results conducted in this study show that the Logistic regression has outperformed all the other models. By the result one can assume that the data is linearly separable and hence logistic regression is giving best results.