Silas implements novel tree-based learning algorithms that yield high predictive performance when dealing with structured data.
Silas adopts various high-performance computing techniques to ensure that the computation is fast and memory-efficient. Silas is built to handle big data and it can be deployed on high performance clusters.
Silas uses formal verification techniques to mathematically verify that the prediction model satisfies user specifications. Further, user specifications can be enforced during the training phase.
Silas adopts the latest automated reasoning techniques to reason about the predictive model and provides insights on the rationale behind the decision-making.
Technical Overview Slides.
Journal Paper. Preprint. BibTex.