CY2D9 Pattern Recognition Coursework Part 1
1. Report the experiments.
a. Download par.m from http://www.personal.reading.ac.uk/~sis01xh/...
Replace line 2 as
c = [1 2 3 4 5], then run par.m.
b. Discuss the shape of the figure produced in relation to the values in c.
c. Repeat the above for c = [-1 -1 0 0 1] and c = [-1 0.5 -1 0 0.5].
2. Suppose that we are given a data set consisting of points xi, j from two classes respectively,
where j =1, 2, denotes class label, and i denotes the data index. The data set is as follows:
Class 1: {1, 2, 0.5}
Class 2: {2, 3, 3.5}
a. Determine the class label for a new data point x = 1.5 using a probabilistic neural network,
with the Gaussian function as window function and σ = 1.
b. How do you find the classification decision boundary of the probabilistic neural network
used in (a) ?
3. Download pnn2D.m from http://www.personal.reading.ac.uk/~sis01xh .., run the code and
describe the results given by the resulting figures, in relation to the data set.