The innovation comes from flexibly generating design matrices and manipulating them to form multiple models easily. This milestone allows

• easily convert your raw data into internal format
• specify how to convert data into a design matrix (design specification)
• build design matrix
• handle indices for each covariate and manipulate them
• reconstruct temporal kernels and initialize weights

At this stage, we will use MATAB's glmfit to do maximum likelihood estimation.