BMEB W4020 Computational Neuroscience: Circuits in the Brain


Course Benefits

    Provides a straighforward theoretical foundation to computational neuroscience.
    Focuses on the intuitive understanding of
(i) Biophysics of Computation,
(ii) Encoding with Neural Circuits, and
(iii) Functional Identification of Neural Circuits.
    Enables the further exploration of key concepts in theoretical neuroscience.

Professor Lazar

    Interests in Computational, Systems and Theoretical Neuroscience
In Silico: Neural Computing Engines and NeuroInformation Processing Machines.
In Vivo: Reverse Engineering the Fruit Fly Brain.
In Silico: The Digital Fruit Fly Brain and Fruit Fly Brain Observatory.
    Further information about the instructor is available under URL: http://www.ee.columbia.edu/~aurel.

Applicable Degree Programs

Most courses 4000-level and above can be credited to all degree programs. All courses are subject to advisor approval.


Lecturer: Professor Aurel A. Lazar
  Office hours: Mondays, 4:00 PM - 6:00 PM, EST, Room 819 CEPSR
  E-mail address: aurel "at" ee.columbia.edu
  Class Web Site: Offered by CourseWorks
TA: Konstantinos Psychas
  Office hours: Fridays, 4:00 PM - 5:30 PM, EST, @EE Student Lounge (1300 Mudd)
  E-mail address: kp2547 "at" columbia.edu
Day and Time: Tuesdays, 7:00 PM - 9:30 PM
Class Location: 633 Seeley W. Mudd Building
Credits for course: 3 points
Prerequisites ELEN E3801 Signals and Systems or Biology W3004 Neurobiology I: Cellular and Molecular Biology plus (Matlab or Python) or the instructor's approval. Biology students who do not have strong prior programming experience should consider taking COMS W1005 Introduction to Computer Science and Programming in MATLAB or ENGI E1006 Introduction to Computing for Engineers and Applied Scientists.
Description: The Biophysics of Computation
Modeling Biological Neurons, The Hudgkin-Huxley Neuron, Modeling Channel Conductances and Synapses as Memristive Systems, I/O Equivalence and Spiking Neuron Models, Bursting Neurons and Central Pattern Generators.
Encoding with Neural Circuits
Stimulus Representation with Time Encoding Machines, Geometry of Time Encoding, Encoding with Neural Circuits with Feedback, Spatio-Temporal Receptive Fields, Population Audio and Video Time Encoding Machines.
Functional Identification of Neural Circuits
Modeling Dendritic Stimulus Processors, Channel Identification Machines, A Fundamental Duality between Neural Decoding and Functional Identification, Identifying Spatio-Temporal Receptive Fields and Biophysical Spike Generators.
Projects in Matlab or Python

[1] Biophysics of Computation Project, [2] Encoding with Neural Circuits Project, and [3] Functional Identification of Neural Circuits Project.
Recommended text(s): Course Notes will be made available.
Reference text(s): ---
Homework(s): 6, mostly writing or adapting simple Matlab code. Students with strong background in programming are encouraged to write code in Python or PyCUDA (instructor's permission required).
Paper(s): ---
Project(s) 3 major projects (see above)
Midterm exam: ---
Final Exam: ---
Grading: The homework grade is based on the average of the best 5 out of 6 homeworks. The final grade has three components: homework and the best two among project [1], project [2] and project [3]. Weighting: 1/5 homework, 2/5 for each of the best 2 out of 3 projects.
Hardware requirements: Laptop for demos
Software requirements: Matlab
Homework submission: Mondays at noon - strict deadline.