Maseeh Summer Undergraduate Research Experience
Funding Source
Frank&Ken (Frankwell Lin and Ken Li) and PSU ECE
Advisor
Christof Teuscher
Document Type
Poster
Publication Date
2025
Abstract
Neural networks are computational models inspired by biological neurons. In this project, recurrent neural networks (RNNs) and spiking recurrent neural networks (SRNNs) were used to solve the Santa Fe Trail problem, a classic machine learning challenge. The goal of this project was to compare how these two kinds of neural networks learned in a reinforcement learning environment and how they performed when evaluated on classic trails. SRNNs not only perform better on trail-solving tasks than traditional RNNs, but they may potentially be more energy efficient. The event driven nature of spiking neural networks enables more efficient computation on neuromorphic hardware, which is promising for cases such as edge computing where power consumption is a large constraint.
Persistent Identifier
https://archives.pdx.edu/ds/psu/45294
Citation Details
McKenna, Kieran, "Solving Trail-Following Tasks Using Recurrent and Spiking Recurrent Neural Networks in a Reinforcement Learning Environment" (2025). Maseeh Summer Undergraduate Research Experience. 94.
https://archives.pdx.edu/ds/psu/45294