[Recommended] 2 Attributes Per Pulse
Consider the attached Ionosphere dataset from the UCI machine learning repository: ‘
‘http://archive.ics.uci.edu/ml/datasets/Ionosphere. The system consists of a phased array of 16 high-frequency antennas with a total transmitted power on the order of 6.4 kilowatts. Received signals were processed using an autocorrelation function whose arguments are the time of a pulse and the pulse number. There were 17 pulse numbers for the system. Instances in this database are described by 2 attributes per pulse number, corresponding to the complex values returned by the function resulting from the complex electromagnetic signal. The targets were free electrons in the ionosphere. “Good” (g) radar returns are those showing evidence of some type of structure in the ionosphere. “Bad” (b) returns are those that do not; their signals pass through the ionosphere.
X is a 351×34 real-valued matrix of predictors. Y is a categorical response: “b” for bad radar returns and “g” for good radar returns. This is a binary classification problem.
Use 5-fold cross-validation to evaluate the classification performance of an LDA and a QDA classifier. Describe which classifier gives you the best performance. Provide the confusion matrix, sensitivity, specificity, total accuracy, F1-score, Roc curve, and area under curve.