Research, software, data and methods. Inspect the context, compare your options, and take a useful next step.
UCI Machine Learning Repository
Vehicle specifications and fuel-use observations for auditing small regression benchmarks.
Investigate Does a model-year holdout change fuel-economy error relative to a random split?
UCI Machine Learning Repository
Generated learning problems for examining how benchmark construction affects classifier comparisons.
Investigate How consistent are simple classifier rankings across a fixed selection of AutoUniv problem settings?
UCI Machine Learning Repository
Wireless localization simulation descriptors for examining error prediction across configurations.
Investigate How does localization-error prediction change when network configuration families are withheld?
UCI Machine Learning Repository
A rule-generated balance problem that separates memorization from learning a physical decision rule.
Investigate Can a learner recover balance outcomes on held-out weight-distance combinations?
UCI Machine Learning Repository
Image-derived banknote features for a compact binary classification and robustness exercise.
Investigate Does banknote classification remain stable under plausible feature scaling and controlled perturbations?
Upstream project repository
Bayesian optimization for bounded black-box objective functions.
Investigate Does a fixed acquisition strategy reduce regret faster than a prespecified random-search baseline on a bounded objective?
UCI Machine Learning Repository
Air-quality measurements from multiple monitoring stations for auditing spatiotemporal forecasting.
Investigate How much does station holdout change particulate-matter prediction compared with chronological holdout alone?
UCI Machine Learning Repository
Forecast-model outputs and weather observations for investigating temperature bias correction.
Investigate Does bias correction improve held-out station and season errors over the raw temperature forecast?
Upstream project repository
Evaluation harness for code language models with task-specific generation and execution workflows.
Investigate Can a chosen task adapter preserve generation settings, sample identity and execution status through evaluation?
UCI Machine Learning Repository
Aggregate bicycle-rental counts and contextual covariates for auditing demand forecasts.
Investigate Does a calendar-aware bike-demand baseline outperform persistence on the final chronological holdout?
UCI Machine Learning Repository
Machining measurements for studying industrial condition recognition under grouped acquisition.
Investigate How does machining-state classification change when machine or recording runs are held out?
Upstream project repository
Model interpretability algorithms for checking attribution behavior against simple functions with known contributions.
Investigate Does a selected attribution method recover known contributions for a linear model under a fixed baseline?
Recorded license evidence is scoped to each source; it is not a blanket permission or a verification result.