This preview shows pages 1–2. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: Assignment (CSE6328 W12) Due: in class on Feb. 9, 2012 You have to work individually. Hand in a hardcopy of your answers before the dead line. No late submission will be accepted. No handwritting is accepted. Direct your queries to Hui Jiang (hj@cse.yorku.ca). 1. Assume we have a random vector x = x 1 x 2 which follows a bivariate Gaus sian distribution: N ( x  , ), where = 1 2 is the mean vector and = 2 1 1 2 1 2 2 2 is the covariance matrix. Derive the formula to compute mu tual information between x 1 and x 2 , i.e., I ( x 1 ,x 2 ). Hints: Refer to the related sections in the reading assignment [W2]. Note there is a mistake in [W2]: equation (99) should be ( n + 1) = n ( n ) . 2. Assume we have two Gaussian distributions: N ( x  1 , 2 1 ) and N ( x  2 , 2 2 ), where 1 and 2 are their means, and 2 1 and 2 2 are their variances. Derive the formula to computer the KL divergence between these two Gaussian distribution.to computer the KL divergence between these two Gaussian distribution....
View
Full
Document
 Winter '12
 Kotakoski

Click to edit the document details