Goto

Collaborating Authors

 South America




OnComputingProbabilisticExplanations forDecisionTrees

Neural Information Processing Systems

For such a reason, the community has started tostudy their probabilistic counterpart, inwhich one requires that the probability ofT(z) = T(x)must beatleast some valueδ (0,1],wherez isa random instance that is compatible withy.








Learning Macroscopic Brain Connectomes via Group-Sparse Factorization

Neural Information Processing Systems

A fundamental challenge in neuroscience is to estimate the structure of white matter connectivity inthehuman brainorconnectomes [14,29]. Connectomes aremadeupofneuronal axonbundles wrapped with myelin sheaths, called fascicles, and connect different areas ofthe brain.