-
Notifications
You must be signed in to change notification settings - Fork 0
jacobsee/Aurora-Perceptron
Folders and files
Name | Name | Last commit message | Last commit date | |
---|---|---|---|---|
Repository files navigation
Aurora Second Assignment for Dr. Cenek's Machine Learning course at UAA This project is written in Go, which can be downloaded at https://golang.org/dl/. It requires the packages go-chart, gonum, and uiprogress. `go get -u github.com/wcharczuk/go-chart`. `go get -u gonum.org/v1/gonum/...` `go get -v github.com/gosuri/uiprogress` This project can be compiled with `go build -o Aurora`. Once compiled, it can be ran with `./Aurora prepare` to generate the histogram file, `histograms.dat`. (takes about 1 minute on my machine) It expects a folder 'images' in the same directory, with subfolders 'aurora' and 'noaurora' containing images to generate histograms for. Then, it can be ran with `./Aurora run histograms.dat 1000 0.01` to train the perceptron for 1000 epochs with a learning rate of 0.01 and run validation against 20% of the data. (takes about 10 seconds on my machine) Demonstration of training: https://asciinema.org/a/9GZrb8fyY0nC2bjXcRnwOV92Y
About
A perceptron for classifying aurora images for Dr. Cenek's Machine Learning course.
Resources
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published