đ¤ Machine Learning
CSharpNumerics includes a lightweight, fully numerical machine learning framework designed for research, experimentation, and educational use. The focus is on transparency, mathematical clarity, and pipeline-based model evaluation â not black-box automation.
namespace CSharpNumerics.ML
| Module | Description |
|---|---|
| âī¸ Supervised AutoML | Automated pipeline search, model selection, and supervised workflow evaluation |
| đ Cross-Validation | K-fold, stratified, and custom validation strategies |
| đˇī¸ Classification | Classification models, decision boundaries, and evaluation workflows |
| đ Regression | Linear, polynomial, and multivariate regression models |
| đ§ Sequence Models | CNN1D, LSTM, Bi-LSTM, sequence layers, and windowed time-series workflows |
| đ Unsupervised AutoML | Automated unsupervised pipelines, clustering search, and experiment workflows |
| đĢ§ Clustering | Clustering algorithms, metrics, and evaluation tools |
| đ˛ Uncertainty Estimation | Bootstrap methods, consensus analysis, and prediction stability estimation |
| đī¸ Dimensionality Reduction | PCA, projection methods, and unsupervised feature preprocessing |
| â Reinforcement AutoML | Reinforcement-learning experiment search, tuning, and evaluation workflows |
| đšī¸ RL Algorithms | Agents, policies, environments, replay buffers, and training diagnostics |