model estimation using training data

I want to estimate a linear model given training data. the linear model equation is given by y=Xb, where 'b' is the model. According to regression definition the linear model is estimated is estimated using the following equation: b = (X'X)X'y

here is the code for my training data mu=zeros(1,10); sigma=eye(10); xTrain = mvnrnd(mu,sigma,1000)'; xTrainT = (xTrain)'; yTrain = randn(1000,1);

now i want to implement the model estimation equation and estimate 'b'.

i used the following code to do that:

betaHat = ((pinv(xTrainT*xTrain))*xTrainT); i get an error saying 'matrix dimension must agree'. i don't know how to solve it everything seems to be fine and the estimation equation for 'b' is also correct.

your help is much appreciated, i am very frustrated thanks