ECE 4363 / ECE 5363 Project 3 solution

Description

Pattern Recognition
1. Synthesize a 2-class problem with samples drawn from two Gaussian distributions with the
following parameters.
m=[zeros(5,1) ones(5,1)];
S(:,:,1)=[0.8 0.2 0.1 0.05 0.01;
0.2 0.7 0.1 0.03 0.02;
0.1 0.1 0.8 0.02 0.01;
0.05 0.03 0.02 0.9 0.01;
0.01 0.02 0.01 0.01 0.8];
S(:,:,2)=[0.9 0.1 0.05 0.02 0.01;
0.1 0.8 0.1 0.02 0.02;
0.05 0.1 0.7 0.02 0.01;
0.02 0.02 0.02 0.6 0.02;
0.01 0.02 0.01 0.02 0.7];
P=[1/2 1/2]’;
2. Draw 𝑁𝑁𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 = 100 training samples with rng(0) and 𝑁𝑁𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 = 10,000 test samples with
rng(100).
3. Use the training samples to learn three classifiers:
i. Naïve Bayes classifier
ii. Bayes classifier that uses MLE for parameter estimation
iii. Bayes classifier that uses the true parameter values
4. Report test errors for each classifier using the test samples
5. Repeat steps 2-4 with 𝑁𝑁𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 = 1000.
Upload a single m-file to Blackboard before midnight on 04/05/19.