Hellebore Machine Learning
Hellebore staff possess a broad background in machine learning and data mining for large data sets. Our experience encompasses supervised learning (classification and regression), unsupervised learning (clustering analysis and dimensionality reduction), and artificial neural networks. Our specific areas of expertise include generalized linear modeling (GLM), principal/independent component analysis (PCA/ICA), and self-organizing maps (SOM).
To augment our machine learning capabilities, we can engineer the modeling and simulation of large-scale systems of differential equations. From traditional methods to leveraging finite difference modeling of reaction-diffusion kinetics in neurons, Hellebore is at the leading edge. From there, our solutions cover foundational concepts in numerical computing (e.g. stability and convergence of numeric solvers, preconditioning) and high-performance computing in distributed and shared-memory clusters.
Beyond scientific simulation, we can build custom 2D and 3D image processing algorithms. We can leverage automated and semi-automated techniques for co-registration, segmentation, and visualization. Our bioinformatics expertise encompasses the processing and statistical analysis of genomic sequencing data.