Neurally-inspired Computing
Add to your list(s)
Download to your calendar using vCal
If you have a question about this talk, please contact Emli-Mari Nel.
This work-in-progress talk is about performing useful computation in a manner that is robust to both the failure of hardware and the re-allocation of hardware to other tasks. The computation is carried out with many identical computational sub-units. Each sub-unit is able to perform some relatively simple computations and has a single bit of storage associated to it.
We deal with three tasks: representing a set, creating an error-correcting code and an associative memory. All solutions have the property that at any point one can take computational sub-units away from the task at hand and the probability of error will increase gracefully. For the error-correcting code and the associative memory one can add sub-units to a task, without changing any of the other units and after a short time they will be used in the task and lower the rate of bit errors.
This talk is part of the Inference Group series.
This talk is included in these lists:
Note that ex-directory lists are not shown.
|