[SOLVED] CS Hive python decision tree NEW YORK INSTITUTE OF TECHNOLOGY

NEW YORK INSTITUTE OF TECHNOLOGY
CSCI 755: Artificial Intelligence 1 (Spring 2021)

You are required to use three classifiers: (1) Nave Bayes, (2) ID3 Decision Trees, and (3) Random Forest, to classify discrete-valued data instances. Use either WEKA software (http://www.cs.waikato.ac.nz/ml/weka/) or Python SciKit (https://scikit-learn.org) to perform classification on the data discussed below.
Data: I posted three files under Project Assignment 1, Handouts, Blackboard. The file car.names describes the data. The file car.data contains the data. This dataset has four classes. The last attribute in car.data is the class label. The file car.c45-names contains descriptions of the class and feature values. Additional information on this dataset can be found on UCI Machine Learning website (http://archive.ics.uci.edu/ml/datasets/Car+Evaluation). When using WEKA, it might be easier to convert the data in car.data into ARFF format (https://www.cs.waikato.ac.nz/ml/weka/arff.html).
Classification Task: Train the classifiers using 60 percent of instances from each class in the car.data file. Test the trained classifiers using the remaining (40 percent) instances in car.data file. Feel free to create separate training and testing data files. Have your own strategy to deal with missing feature values (e.g., remove instances with missing features or fill in the missing feature values with the most popular value.).
Program Metrics: Report the classification accuracy, per class classification accuracy, and confusion matrix on the test instances
Deliverables:
A well-written report containing comparisons of the results of the classifiers. Use classification accuracy (# of instances correctly classified/total # of instances presented for classification), per class classification accuracy, and confusion matrix to compare the results. You should be able to demonstrate your results. In your report, use screenshots and visualization wherever possible to substantiate/prove your results.[60 points]
Compare the accuracies of the classifiers when trained on 50 percent, 75 percent, and 90 percent of instances per class (remaining should be used for testing). In the report, present an analysis on how the amount of training data impacts classification accuracy for each classifier (which classifier is more sensitive to training sample size)?. [40 points]