No user-submitted related servers found.
Scored across 9 tools
Each tool has a clearly distinct purpose: train, single-model predict (LR vs KNN), combined predict, metrics, model selection, SHAP explanation, sample input, and audit export. The only mild overlap is predict_both vs the two single-model predict tools, but predict_both is explicitly a combined convenience wrapper and is easy to tell apart.
Nearly all names are consistent snake_case, with most following a verb_noun pattern (train_models, predict_knn, select_best_model, export_parquet_audit). Minor deviations exist: confusion_matrix_metrics is noun-only and shap_explain_selected uses a domain-prefixed verb form.
Nine tools map cleanly onto the ML pipeline stages (train, predict, evaluate, select, explain, export) without excessive granularity. Each tool earns its place with no redundant filler.
The surface covers the full lifecycle: training, individual and combined prediction, evaluation metrics, model selection, explainability, plus audit export and a sample-input helper. Minor gaps remain—no batch prediction or data preprocessing/loading step—but core workflows are fully served.