Project Title
Preprocessing with Clustering to Avoid Convolution
Document Type
Event
Start Date
10-5-2019 3:30 PM
End Date
10-5-2019 6:30 PM
Description
Convolutional Neural-Networks (CNN) are used extensively for machine learning. However, CNNs are expensive; we investigate an approach of preprocessing images such that we can avoid convolution. We use this approach to classify handwritten digits, still using a Neural-Network architecture.
Discipline
Math
Research Mentor(s)
Wai Lau, Sarah McCord
Copyright Status
http://rightsstatements.org/vocab/InC/1.0/
Additional Rights Information
Copyright held by author(s).
Preprocessing with Clustering to Avoid Convolution
Convolutional Neural-Networks (CNN) are used extensively for machine learning. However, CNNs are expensive; we investigate an approach of preprocessing images such that we can avoid convolution. We use this approach to classify handwritten digits, still using a Neural-Network architecture.