menu

Learning circuits for memristor-based neuromorphic circuits

In this workgroup, I propose to work on algorithmic and circuit design ideas for designing memristor-based neuromorphic designs. Some of the interesting problems to be considered could be variability-tolerant algorithms and circuits, synapse design, synaptic learning rules, etc.

Login to become a member send

Timetable

Day Time Location
Thu, 28.04.2016 14:00 - 15:00 Panorama Room
Mon, 02.05.2016 14:30 - 15:30 Panorama room
Wed, 04.05.2016 14:00 - 15:00 Panorama Room
Icon Screen shots of discussion sheets (2.7 MB)

Algorithms and circuits discussed

1. Metaplasticity

2. Low precision forward pass and high precision backward pass - Gradient Descent

3. Oscillator synchronization

4. Target Propagation

5. Pulse-based STDP

6. Memory trace, SOM, WTA

7. Restricted Boltzmann Machines

8. Spiking Deep Networks

 

Leader

Manu Nair

Members

Alessandro AImar
Christopher Bennett
Enrico Calabrese
Gabriel Andres Fonseca Guerra
Marina Ignatov
Giacomo Indiveri
bernabe linares-barranco
Bragi Lovetrue
Lorenz Muller
Manu Nair
Melika Payvand
Christian Pehle
Francesca Puppo
Ning Qiao
Yulia Sandamirskaya
Jaak Simm
Evangelos Stromatias
Nikolaos Vasileiadis
Borys Wrobel
Qi Xu