philsaurabh/parkinson-disease-ensemble-learning

To implement an Ensemble Classifier of Logistic regression, Naive Bayes & Decision Tree. The goal is to implement a classification model to predict the “Status” feature in the provided Parkinson dataset.

Parkinson-Disease-Ensemble-Learning

To implement an Ensemble Classifier of Logistic regression, Naive Bayes & Decision Tree. The Goal is to implement a classification model to predict the “Status” feature in the provided Parkinson dataset.(Excluded the ID column). Constructed the following -

1. Ensemble using Majority voting ;
2. Ensemble using weighted voting; accuracy of the classifiers on the development set (20% of the overall data) is to be used as the weight. A. For each of the ensembles compare the performance of the ensemble to each of the base models. Is there always a benefit of ensembles? B. Redoing ensemble with 5 versions of each of the specified classifiers(each version has different hyperparameters, if possible, for ex. For DT different versions can be generated by varying tree depth and branching factor, for logistic regression you can make variations by setting regularization etc.). How does this ensemble compare to the one with only one of each.(Answers in .docx file)