Q1. In this problem, you will explore classification based on discriminant analysis and MLE.[4] a…

Q1. In this problem, you will explore classification based on discriminant analysis and MLE.[4] a…
 Q1. In this problem, you will explore classification based on discriminant analysis and MLE.[4] a. Generate 200 multivariate (dimension = 2) Bernoulli distributed samples, 100 of those samples (Class 1) should be generated from (u1=[ 0.5 0.8] ) and the other 100 (Class2) should be generated from (u2=[0.9 0.2]). Assume conditional independence. You can use any inbuilt library. b. Select 50 samples out of 100 samples from class 1, call them training set and use those samples to compute the class conditional parameters => ul. The remaining 50 samples will be used later for classification, call them test set. – You need to compute using MLE. You cannot use an inbuilt function that gives MLE. – Now compute MLE using only n samples at a time where n = 1,2,3,……50. Analyze MLE vs n plot and report your observations. c. Repeat step (b) for class 2 samples. d. Plot training samples using a scatter plot. May 16 2022 05:47 PM