AutoData_Statistical_Analysis-and-Voting-by-Discrimination

This project contains statistical data analysis of Auto dataset which is about different cars and various features of cars. Basic analysis were done using mean, standard deviation, scatterplot, histograms, quantile plot, correlation. Advanced analysis includes separation of HIGHmpg and LOWmpg using quantile plot. Then, assigning score to each case and using t-test formula to calculate discriminatory power. After that, threshold value was computed and prediction was done through voting mechanism for each case. Lastly, confusion matrix was generated to see global accuracies for train and test set.