We consider the problem of distinguishing between two groups of subjects based on their white matter connectivity. Our goal is to discriminate between the two groups (the general classification problem in machine learning) not based on global features such as the number of fibers, degree distribution, distribution of streamlines, centrality measures, shortest path distances, but on spatially localized differences in white matter connectivity. Localizing the analysis can reveal important elements that may be lost in global connectivity measures. Spatial localization of differences is also readily amenable to interpretation and clinical diagnosis. A method that separates between two groups based on local white matter connectivity has to overcome two challenges. The first is that of coping with natural variation between members of a group (even repeated imaging on the same subject does not produce identical streamlines) to obtain a significant separation across groups. The second challenge is that of coping with large amounts of data so that the two groups to be compared can be chosen on demand based on attributes such as age, gender, state of disease, etc.