This dataset include two different sets of images: Class A and Class B. These images represent the optical responses of proteins interfaced with liquid crystal. Two type of proteins are placed on liquid crystals interfaces to trigger different responses. The dataset given here is used to discriminate between the reponses and analyze them.
FOR COURSE USE ONLY - By downloading these files, you agree to the following terms: These files are for use in the course; not for any other type of use. Please do not post, redistribute, share, or offer these downloads as your own or with credit. Please do not distribute these files to others in the course, but refer others here to download for themselves instead. Thank you for honoring this agreement! Please contact Abdulelah Alshehri or Professor Fengqi You for any question.
In this project, your goal is to correctly classify images from the dataset. We encourage you to experiment with different algorithms to learn first-hand what works for the dataset and how different machine learning techniques compare.
Train (use for training and validation)
- A: 962
- B: 628
Please use the same dimensions (165 x 165) as the provided images as an input to your model.
- Computer vision including convolutional neural networks
- Classification and Clustering methods such as SVM, KNN, and T-SNE
You should develop a model that correctly provide the correct label for each image in the test set.
This project is evaluated on the accuracy of your predictions (the percentage of images you correctly label). The perfromance attained on the test set (hidden here) should be at least 90% to get the full credit.
Don't forget to cross-validate the performance of your model!
- Michael Tsuei
- Professor Nicholas Abbott
- Professor Anthony Hay
Copyright © PEESE @ Cornell University
All rights reserved