Structured data extraction with LLM majority vote
VoteEnsemble: Ensemble methods for machine/deep learning and stochastic programming with guaranteed generalization.
Data Science Tools, Libraries, Classes and Functions
Equal-cost LLM strategy benchmark: does independent sampling beat self-critique loops? Harness + live dashboard (Ollama / OpenRouter).
HYPO_RFS is an algorithm for performing exhaustive grid-search approach for tuning the hyper-parameters of Ranking Feature Selection (RFS) approaches.
Zero-shot LLM text classification with calibrated confidence via k independent calls and hard majority voting. No training, no logits required.
To associate your repository with the majority-vote topic, visit your repo's landing page and select "manage topics."