PulseAugur
LIVE 03:32:03
research · [2 sources] ·
0
research

SOC-ICNN: From Polyhedral to Conic Geometry for Learning Convex Surrogate Functions

Researchers have introduced SOC-ICNN, a novel neural network architecture that expands the representational capabilities beyond classical ReLU-based Input Convex Neural Networks (ICNNs). By generalizing from Linear Programming (LP) to Second-Order Cone Programming (SOCP), SOC-ICNNs can natively incorporate smooth curvature while maintaining an optimization-theoretic interpretation. This advancement broadens the representational space without increasing computational complexity, leading to improved function approximation and decision-making quality in downstream tasks. AI

Summary written by None from 2 sources. How we write summaries →

IMPACT Expands the representational capacity of convex neural networks, potentially improving performance in optimization-related AI tasks.

RANK_REASON Academic paper introducing a new neural network architecture with theoretical and experimental improvements.

Read on arXiv stat.ML →

COVERAGE [2]

  1. arXiv stat.ML TIER_1 · Kang Liu, Jianchen Hu ·

    SOC-ICNN: From Polyhedral to Conic Geometry for Learning Convex Surrogate Functions

    arXiv:2604.22355v1 Announce Type: cross Abstract: Classical ReLU-based Input Convex Neural Networks (ICNNs) are equivalent to the optimal value functions of Linear Programming (LP). This intrinsic structural equivalence restricts their representational capacity to piecewise-linea…

  2. arXiv stat.ML TIER_1 · Jianchen Hu ·

    SOC-ICNN: From Polyhedral to Conic Geometry for Learning Convex Surrogate Functions

    Classical ReLU-based Input Convex Neural Networks (ICNNs) are equivalent to the optimal value functions of Linear Programming (LP). This intrinsic structural equivalence restricts their representational capacity to piecewise-linear polyhedral functions. To overcome this represent…