cacc implementation function: https://www.mathworks.com/matlabcentral/fileexchange/41740-discretization-methods-class-attribute-contingency-coefficient-cacc-matlab
Assignment: The assignment will use the Iris Data Set (attached as text file) and which can also be found at: http://archive.ics.uci.edu/ml/datasets/Iris
This assignment has two parts:
Part 1. Discretization of continuous valued attributes:
A) The first approach is to use a naive discretization by just round off the decimal values, so that all the attributes are integer-valued.
B) The second approach is to Run the Class-Attribute Contingency Coefficient (CACC) discretization algorithm
The paper describing the algorithm is attached to this assignment.
A MATLAB function implementing it can be found at:
Part 2. For each discretization approach, use the discretized attributes to implement ID3 in Matlab (if you use code lifted from some online source please document that as a comment in your program and reference in the report. Provide at least 5 runs (different training and test sets), and the corresponding accuracies. You can also use the MATLAB function confusion which will generate a confusion matrix.