diff --git a/common/gsog.m b/common/gsog.m
index 037c2fc..3a2b5b5 100644
--- a/common/gsog.m
+++ b/common/gsog.m
@@ -1,5 +1,6 @@
 function [Q, R] = gsog(X)
 % Gram-Schmidt orthogonalization
+% Written by Mo Chen (sth4nth@gmail.com).
 [d,n] = size(X);
 m = min(d,n);
 R = eye(m,n);
diff --git a/common/gson.m b/common/gson.m
index 8ff6683..1fcdda4 100644
--- a/common/gson.m
+++ b/common/gson.m
@@ -1,5 +1,6 @@
 function [Q, R] = gson(X)
 % Gram-Schmidt orthonormalization which produces the same result as [Q,R]=qr(X,0)
+% Written by Mo Chen (sth4nth@gmail.com).
 [d,n] = size(X);
 m = min(d,n);
 R = zeros(m,n);
diff --git a/common/invpd.m b/common/invpd.m
index c8ab2dd..572666c 100644
--- a/common/invpd.m
+++ b/common/invpd.m
@@ -1,6 +1,7 @@
 function W = invpd(M)
 % Compute A\B where A is a positive definite matrix
-%   A: a positive difinie matrix
+% Input:
+%   M: a positive difinie matrix
 % Written by Michael Chen (sth4nth@gmail.com).
 [U,p] = chol(M);
 if p > 0
diff --git a/common/log1pexp.m b/common/log1pexp.m
index 3b5beb2..7ad0b9d 100644
--- a/common/log1pexp.m
+++ b/common/log1pexp.m
@@ -1,5 +1,5 @@
 function y = log1pexp(x)
-% accurately compute y = log(1+exp(x))
+% Accurately compute y = log(1+exp(x))
 % reference: Accurately Computing log(1-exp(|a|)) Martin Machler
 seed = 33.3;
 y = x;
diff --git a/common/logsumexp.m b/common/logsumexp.m
index e622ef4..9838e4f 100644
--- a/common/logsumexp.m
+++ b/common/logsumexp.m
@@ -1,7 +1,7 @@
 function s = logsumexp(X, dim)
 % Compute log(sum(exp(X),dim)) while avoiding numerical underflow.
 %   By default dim = 1 (columns).
-% Written by Michael Chen (sth4nth@gmail.com).
+% Written by Mo Chen (sth4nth@gmail.com).
 if nargin == 1, 
     % Determine which dimension sum will use
     dim = find(size(X)~=1,1);
diff --git a/common/mgsog.m b/common/mgsog.m
index 9a6cab6..003ce87 100644
--- a/common/mgsog.m
+++ b/common/mgsog.m
@@ -1,5 +1,6 @@
 function [Q, R] = mgsog(X)
 % Modified Gram-Schmidt orthogonalization
+% Written by Mo Chen (sth4nth@gmail.com).
 [d,n] = size(X);
 m = min(d,n);
 R = eye(m,n);
diff --git a/common/mgson.m b/common/mgson.m
index 3550f13..ecea8b3 100644
--- a/common/mgson.m
+++ b/common/mgson.m
@@ -1,6 +1,7 @@
 function [Q, R] = mgson(X)
 % Modified Gram-Schmidt orthonormalization (numerical stable version of Gram-Schmidt algorithm) 
 % which produces the same result as [Q,R]=qr(X,0)
+% Written by Mo Chen (sth4nth@gmail.com).
 [d,n] = size(X);
 m = min(d,n);
 R = zeros(m,n);
diff --git a/common/solvpd.m b/common/solvpd.m
index 198836d..01000a6 100644
--- a/common/solvpd.m
+++ b/common/solvpd.m
@@ -1,7 +1,7 @@
 function V = solvpd(A,B)
 % Compute A\B where A is a positive definite matrix
 %   A: a positive difinie matrix
-% Written by Michael Chen (sth4nth@gmail.com).
+% Written by Mo Chen (sth4nth@gmail.com).
 [U,p] = chol(A);
 if p > 0
     error('ERROR: the matrix is not positive definite.');
diff --git a/common/standardize.m b/common/standardize.m
index 11a5805..233fadd 100644
--- a/common/standardize.m
+++ b/common/standardize.m
@@ -1,7 +1,7 @@
 function [Y, s] = standardize(X)
 % Unitize the vectors to be unit length
 %   By default dim = 1 (columns).
-% Written by Michael Chen (sth4nth@gmail.com).
+% Written by Mo Chen (sth4nth@gmail.com).
 if nargin == 1, 
     % Determine which dimension sum will use
     dim = find(size(X)~=1,1);
diff --git a/common/symeig.m b/common/symeig.m
index c029259..6800a45 100644
--- a/common/symeig.m
+++ b/common/symeig.m
@@ -2,7 +2,7 @@
 % Compute eigenvalues and eigenvectors of symmetric matrix
 %   m == 's' smallest (default)
 %   m == 'l' largest
-% Written by Michael Chen (sth4nth@gmail.com).
+% Written by Mo Chen (sth4nth@gmail.com).
 if nargin == 2
     m = 's';
 end
diff --git a/common/ud.m b/common/ud.m
index 0eb83e0..4f1ea34 100644
--- a/common/ud.m
+++ b/common/ud.m
@@ -1,5 +1,6 @@
 function [U, D] = ud(X)
 % UD factorization U'*D*U=X'*X;
+% Written by Mo Chen (sth4nth@gmail.com).
 [~,R] = qr(X,0);
 d = diag(R);
 D = d.^2;
diff --git a/common/unitize.m b/common/unitize.m
index 4a47556..feb12bb 100644
--- a/common/unitize.m
+++ b/common/unitize.m
@@ -1,7 +1,7 @@
 function [Y, s] = unitize(X, dim)
 % Unitize the vectors to be unit length
 %   By default dim = 1 (columns).
-% Written by Michael Chen (sth4nth@gmail.com).
+% Written by Mo Chen (sth4nth@gmail.com).
 if nargin == 1, 
     % Determine which dimension sum will use
     dim = find(size(X)~=1,1);
diff --git a/other/loggmpdf.m b/other/loggmpdf.m
index a86af95..01c2c25 100644
--- a/other/loggmpdf.m
+++ b/other/loggmpdf.m
@@ -1,5 +1,6 @@
 function r = loggmpdf(X, model)
-
+% Compute log pdf of a Gaussian mixture model.
+% Written by Mo Chen (sth4nth@gmail.com).
 mu = model.mu;
 Sigma = model.Sigma;
 w = model.weight;
diff --git a/other/plotgm.m b/other/plotgm.m
index 02d7ac3..5915b4e 100644
--- a/other/plotgm.m
+++ b/other/plotgm.m
@@ -1,5 +1,6 @@
 function plotgm(X, model)
-% Written by Michael Chen (sth4nth@gmail.com).
+% Plot 2d Gaussian mixture model.
+% Written by Mo Chen (sth4nth@gmail.com).
 level = 64;
 n = 256;
 
diff --git a/other/plotkde.m b/other/plotkde.m
index eaee983..0b51a66 100644
--- a/other/plotkde.m
+++ b/other/plotkde.m
@@ -1,6 +1,6 @@
 function plotkde(X, sigma2)
-% Written by Michael Chen (sth4nth@gmail.com).
-
+% Plot 2d kernel density.
+% Written by Mo Chen (sth4nth@gmail.com).
 if nargin < 2
     sigma2 = 1e-1;
 end