Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Supervised Distributional Reduction via Optimal Transport and Dependence Maximization

About

Learning representations that capture both intrinsic data geometry and target-relevant structure remains a fundamental challenge, particularly in settings where data reduction must balance compression with predictive fidelity. While distributional reduction-encompassing joint clustering and dimensionality reduction-offers a principled way to summarize data, its supervised variants remain relatively under-explored, despite the importance of retaining task-relevant signal for downstream prediction and decision-making. We propose Supervised Distributional Reduction (SDR), an algorithm for learning target-aware representations by combining optimal transport with explicit dependence maximization. SDR builds on the Fused Gromov-Wasserstein (FGW) objective to align the relational structure of the input distribution with a set of representative points, while augmenting it with a direct dependence term that encourages the learned embeddings to capture predictive signal more explicitly. This results in compact representations that reflect both geometric structure and supervision. Beyond representation learning, SDR naturally induces a data-dependent, non-stationary geometry that can be leveraged for settings such as Gaussian Process (GP) modelling. By redefining distances through target-aware distributional alignment, SDR enables the construction of adaptive kernels that respond to local variations in both data geometry and supervision, offering an optimal transport-based perspective on non-stationary kernel design.

Sai-Aakash Ramesh, Archit Sood, Andrew Corbett, Tim Dodwell• 2026

Related benchmarks

TaskDatasetResultRank
ClassificationCOIL-20
Accuracy0.99
96
RegressionUCI ENERGY (test)
Negative Log Likelihood0.37
62
RegressionBoston UCI (test)--
36
RegressionS-curve 100 points (test)
R2 Score0.96
6
RegressionSwiss Roll 100 points (test)
R2 Score97
6
RegressionFriedman 500 points (test)
R291
6
RegressionS-curve 100 points (train)
R20.98
6
RegressionSwiss Roll 100 points (train)
R2 Score0.99
6
RegressionFriedman 500 points (train)
R20.98
6
ClassificationMNIST 2k
MLP Metric-0.23
5
Showing 10 of 10 rows

Other info

Follow for update