LOTTERY: Learning from Reference-Only Samples in Two-Sample Testing under Size Asymmetry
About
Data-adaptive two-sample testing assesses if two samples come from the same distribution, using a discrepancy learned from the data (e.g., via kernel-based feature representations). Such methods typically rely on data splitting to decouple learning from testing and control type I error. However, this paradigm is ill-suited to few-shot settings with severe sample-size imbalance: abundant reference samples are available, while only a handful of query samples arrive. In this paper, we show how this imbalance can be leveraged constructively. Using abundant reference data, we learn reference-dependent representations that summarize salient structure of the reference distribution and provide informative signals for detecting departures. We incorporate a collection of representation families that capture both global and local structure, and adaptively weight them using only reference samples via an uncertainty-guided principle. Theoretically, we establish permutation-based type I error control and show consistency of the aggregated test: as the sample sizes grow, the test power converges to one whenever the representation set contains at least one consistent representation. Empirically, our aggregation achieves strong performance across a range of benchmarks while retaining type I error control.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Two-sample testing | higgs | Test Power100 | 159 | |
| Two-sample testing | CIFAR10-RES18 (test) | Test Power100 | 97 | |
| Two-sample testing | CIFAR10 WRN28 | Test Power66.9 | 49 | |
| Two-sample testing | BLOB (test) | Test Power72.5 | 49 | |
| Two-sample testing | Blob | Test Power1 | 49 | |
| Two-sample testing | CIFAR10-WRN8 | Test Power95.8 | 49 |