Classification model selected

  • Category:
    Logic & Programming
  • Document type:
    Assignment
  • Level:
    Undergraduate
  • Page:
    1
  • Words:
    161

Actual Model Chosen

The choice of the actual is determined by a comparison of the performance of accuracy and the error rates on validation dataset.

Decision Tree

Random Forest

Boosting

Accuracy In %

Error Rate in %

On the basis of accuracy of classification of the data set, Random forest classifier is the most accurate. It has the highest accuracy and hence the most reduced error rate.

Discussion

As an ensemble model, the random forest enhances representation of the target function through an understanding of more subtle relationships of the given data set. The bagging incorporated elps to improve the generalization error by reducing the variance of the base classifiers hence leading to a higher accuracy compared to others. The elimination of pruning leads to a less biased model. This model is also more resistant to outlier and noise impact. Just for an addition, this classifier is faster and it therefore as every aspect worth considering for this particular classification process.