Home

This is the discrete neural nets wiki. On this page we'll keep track of tasks we'd like to perform. Put your name next to something in parenthesis if you'd like to do it. (And talk to us about it of course.)

• Change __getitem__ to __call__ in Operation class. (Still one issue with test_binary_image_train_endomorphisms.py.) (Hussein, Charlotte to fix last issue)
• Make a Relation class which is sparse to replace the binary images. (Lily, Kevin)
• Switch to using MNIST CSV instead of the TensorFlow MNIST so we can drop the dependency. (Kevin)
• Make polymorphisms cache values by default. (Charlotte)
• Add Git workflow to the readme. (Charlotte)
• Remove the path variable setting on all the test scripts. (Charlotte)
• Add a description to the readme. (Charlotte)
• Describe in the readme how to point the test scripts to the src folder for imports. (Charlotte)
• Test higher-arity classification. (TJ, Karam)
• Clean up dominion.py. Make scripts in the tests folder which test the dominion polymorphisms. (Hussein)
• Implement hyperoctahedral automorphisms for higher arity relations. (Lily, Kevin)
• Implement other higher arity versions of Hamming graph polymorphisms. (Lily, Kevin)
• Make the wiki a history of what we've done with check boxes. (Charlotte)
• Make file storage and access faster for dominion polymorphisms.