Image Sorter is a image clustering application.
This is an Academic Project.
This project uses K-means algorithm to cluster random images. Since most of the data out there is not labelled, unsupervised learning methods such as K-means is very useful to see patterns. This algorithm works efficiently when the groups are distinct and non-overlapping. The images for this project was collected from various sources such as kaggle and pyimagesearch.
This project involves exploring various methods to initialize the cluster centroids to see which performs the best.
Since we know the labels for this particular dataset, we can check how the algorithm performed. The last part of this project shows how a supervised learning algorithm such as Neural Networks with a single hidden layer performs on these images.