Increased model stability on unseen data.
Solid understanding of models behavior.
Improved model stability on unseen data.
Stability Against noise
Enhance the generalisation and prediction power of general ML algorithms
Robust performance on small data sets and outliers.
Invariance under deformation
Simplify and transform models to increase robustness
Outstanding visualization features.
Preservation of information related to data proximity and shape
Our high-performance toolkit for topological data analysis: A fully scikit-learn compatible API for persistent homology and the mapper algorithm.
Scalability-oriented topological deep-learning library for seamless integration of Pytorch with state-of-the-art topological algorithms.