Published June 25, 2021
| Submitted
Report
Open
Neural Networks Computations with DOMINATION Functions
- Creators
- Kilic, Kordag Mehmet
- Bruck, Jehoshua
Abstract
We study a new representation of neural networks based on DOMINATION functions. Specifically, we show that a threshold function can be computed by its variables connected via an unweighted bipartite graph to a universal gate computing a DOMINATION function. The DOMINATION function consists of fixed weights that are ascending powers of 2. We derive circuit-size upper and lower bounds for circuits with small weights that compute DOMINATION functions. Interestingly, the circuit-size bounds are dependent on the sparsity of the bipartite graph. In particular, functions with sparsity 1 (like the EQUALITY function) can be implemented by small-size constant-weight circuits.
Attached Files
Submitted - etr151.pdf
Files
etr151.pdf
Files
(256.8 kB)
Name | Size | Download all |
---|---|---|
md5:4276e3b9f386549848157eca04fb65db
|
256.8 kB | Preview Download |
Additional details
- Eprint ID
- 109572
- Resolver ID
- CaltechAUTHORS:20210624-214748102
- Created
-
2021-06-25Created from EPrint's datestamp field
- Updated
-
2021-06-25Created from EPrint's last_modified field
- Caltech groups
- Parallel and Distributed Systems Group
- Series Name
- Parallel and Distributed Systems Group Technical Reports
- Series Volume or Issue Number
- etr151