Category Theory in Machine Learning
arXiv:2106.07032
Abstract
Over the past two decades machine learning has permeated almost every realm of technology. At the same time, many researchers have begun using category theory as a unifying language, facilitating communication between different scientific disciplines. It is therefore unsurprising that there is a burgeoning interest in applying category theory to machine learning. We aim to document the motivations, goals and common themes across these applications. We touch on gradient-based learning, probability, and equivariant learning.
References in corpus (36)
- UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction
- An overview of gradient descent optimization algorithms
- Language Models are Few-Shot Learners
- Graph Neural Network for Traffic Forecasting: A Survey
- On the Generalization of Equivariance and Convolution in Neural Networks to the Action of Compact Groups
- Gauge Equivariant Convolutional Networks and the Icosahedral CNN
- A synthetic approach to Markov kernels, conditional independence and theorems on sufficient statistics
- Disintegration and Bayesian Inversion via String Diagrams
- Interactive Visualization of 2-D Persistence Modules
- Picturing classical and quantum Bayesian inference
- Seven Sketches in Compositionality: An Invitation to Applied Category Theory
- A General Theory of Equivariant CNNs on Homogeneous Spaces
- Natural Graph Networks
- Persistent Clustering and a Theorem of J. Kleinberg
- Representable Markov Categories and Comparison of Statistical Experiments in Categorical Probability
- Convex Spaces I: Definition and Examples
- Profunctor Optics, a Categorical Update
- Open Diagrams via Coend Calculus
- Categories of Optics
- Reverse Derivative Ascent: A Categorical Approach to Learning Boolean Circuits
- What is Applied Category Theory?
- Bayesian Updates Compose Optically
- Bayesian machine learning via category theory
- Functorial Hierarchical Clustering with Overlaps
- Lenses and Learners
- Generalized Lens Categories via functors
- Categorical Stochastic Processes and Likelihood
- Poly: An abundant categorical setting for mode-dependent dynamics
- Consistency constraints for overlapping data clustering
- Compositional Deep Learning
- On Universal Approximation by Neural Networks with Uniform Guarantees on Approximation of Infinite Dimensional Maps
- Limits of bimorphic lenses
- Differentiable Causal Computations via Delayed Trace
- Characterizing the invariances of learning algorithms using category theory
- Categorical Aspects of Parameter Learning
- Flattening Multiparameter Hierarchical Clustering Functors