Course or series
Introduction to Neural Computation
MIT OpenCourseWare · Source-checked annotation
1. Overview and Ionic Currents
Michale Fee begins neural computation with diffusion, concentration gradients and electrical forces as ingredients of neuronal currents.
MIT OpenCourseWare · Source-checked annotation
2. RC Circuit and Nernst Potential
Models neuronal responses using membrane resistance, capacitance and an electrical-circuit analogy, introducing the source of neuronal potentials.
MIT OpenCourseWare · Source-checked annotation
3. Nernst Potential and Integrate and Fire Models
Continues the treatment of neuronal potentials and introduces an integrate-and-fire model.
MIT OpenCourseWare · Source-checked annotation
4. Hodgkin-Huxley Model Part 1
Introduces the Hodgkin–Huxley circuit model through voltage-clamp measurements and the time and voltage dependence of potassium current.
MIT OpenCourseWare · Source-checked annotation
5. Hodgkin-Huxley Model Part 2
Continues the Hodgkin–Huxley model, voltage-clamp analysis and the role of conductance in describing current.
MIT OpenCourseWare · Source-checked annotation
6. Dendrites
Extends circuit modelling to dendrites, spatial dimensions and a two-compartment representation.
MIT OpenCourseWare · Source-checked annotation
7. Synapses
Introduces models of synaptic transmission, saturation and inhibition at different neuronal locations.
MIT OpenCourseWare · Source-checked annotation
8. Spike Trains
Introduces extracellular signals, spike detection, firing rates and stimulus-aligned spike summaries.
MIT OpenCourseWare · Source-checked annotation
9. Receptive Fields
Introduces mathematical descriptions of responses to stimuli, including receptive fields, separability and spike-triggered averaging.
MIT OpenCourseWare · Source-checked annotation
10. Time Series
Connects models of irregular spike trains with the tools used to compare signals over time: convolution, correlation and Fourier series.
MIT OpenCourseWare · Source-checked annotation
11. Spectral Analysis Part 1
Introduces complex Fourier series, continuous and discrete transforms, and power spectra in a neural-computation course.
MIT OpenCourseWare · Source-checked annotation
12. Spectral Analysis Part 2
Moves from Fourier transforms to estimating spectra, filtering signals in the frequency domain and removing line noise.
MIT OpenCourseWare · Source-checked annotation
13. Spectral Analysis Part 3
Reviews spectral estimation, windows, tapers, spectrograms and the trade-off represented by the time-bandwidth product.
MIT OpenCourseWare · Source-checked annotation
14. Rate Models and Perceptrons
Introduces tractable network models through rate descriptions, receptive fields, vector algebra and perceptrons.
MIT OpenCourseWare · Source-checked annotation
15. Matrix Operations
Uses perceptrons and their learning rule to discuss neuronal logic, linear separability, invariance and matrix transformations.
MIT OpenCourseWare · Source-checked annotation
16. Basis Sets
Connects two-layer feed-forward networks with matrix transformations, basis sets, linear independence and changes of basis.
MIT OpenCourseWare · Source-checked annotation
17. Principal Components Analysis
Develops principal components analysis using covariance, distributions, eigenvalues and eigenvectors.
MIT OpenCourseWare · Source-checked annotation
18. Recurrent Networks
Introduces mathematical descriptions of recurrent networks in relation to memory storage and decision-making.
MIT OpenCourseWare · Source-checked annotation
19. Neural Integrators
Describes neural integrator models in relation to short-term memory, accumulated decision evidence, eye movements and persistent network activity.
MIT OpenCourseWare · Source-checked annotation
20. Hopfield Networks
Examines attractor and winner-take-all networks, long-term memory models and Hopfield-network capacity.