Chandan Singh


phd student at uc berkeley exploring the intersection of machine learning and neuroscience


interpretable ml
feature importance
Connectomics , neural coding
“hierarchical interpretations for neural network predictions”, Singh, Murdoch, & Yu, 2017, arXiv “large scale image segmentation with structured loss based deep learning for connectome reconstruction”, Funke, Tschopp, Grisaitis, Sheridan, Singh, Saalfeld, & Turaga, 2017, TPAMI “linearization of excitatory synaptic integration at no extra cost”, Morel, Singh, & Levy, 2018, J. Comp Neuro
  “a constrained, weighted-l1 minimization approach for joint discovery of heterogeneous neural connectivity graphs” - Singh, Wang, & Qi, 2017, NIPS Workshop “a consensus layer V pyramidal neuron can sustain interpulse-interval coding” - Singh & Levy, 2017, Plos One


  • Fall 17 - Present

    Berkeley bin yu lab

    phd research in interpretable ml and neuroscience

  • Worked on unsupervised pretraining of CNNs for semantic segmentation.

  • Fall 16 - Spring 17

    UVA Yanjun Qi Research Lab

    Contributed to development of novel weighted-L1, multi-task Gaussian graphical models

  • Summer 15, Winter 15, Summer 16

    HHMI Srini Turaga Lab

    Contributed to enhanced ML implementations for neural image segmentation

  • Fall 14 - Fall 16

    UVA William Levy Lab

    Simulated stochastic neurons to determine mutual information, variability, energy efficiency, and threshold

  • Simulated extracellular neural recordings via Neurocube Matlab scripts

  • Summer 13 - Spring 14


    Developed web app for simultaneous task coordination + Android app to increase the data storage capacity of QR codes.

notes and more (in progress)