Code and data for machine learning on monomial ideals. Specifically, these files are meant to accompany my doctoral dissertation, Probability and Machine Learning in Combinatorial Commutative Algebra, and that document provides the complete context.
Contents:
- MonLearning.py: Keras/TensorFlow code defining the neural network architecture and implementing training and testing routines.
- TrainingData/: A directory containing the complete training data, organized into subdirectories corresponding to the families of monomial ideals described in Chapter 7.
- HilbertML.m2: Macaulay2 code implementing flexible pivot rule choices for Hilbert series computations, counting base cases, etc.
- IdealML.m2: Miscellaneous M2 code for generating random monomial ideals of various kinds, computing invariants, file I/O and organization for training data, etc.