Stochastic Optimization of Tree Tensor Networks

Marius Willner, Maximilian Scharf, André Uschmajew, Timo Felser, Marco Trenti

arXivarxiv/2609.00870Sep 1, 2026math.OCscore 0

Tensor networks, originally developed for quantum many-body physics, are promising models for machine learning. We derive stochastic Riemannian optimizers for tree tensor networks (TTNs) on both their parameter and quotient manifolds, including adaptive and learning-rate-free schemes suitable for minibatch training. Using a hybrid CNN-TTN architecture, we evaluate the methods on Fashion-MNIST, CIFAR10, and Imagenette. The proposed optimizers achieve predictive performance comparable to unconstrained optimization while enabling numerically stable downstream compression.

SourceJSON

Comments and reviews

No comments yet.

API

curl /api/v1/records/arxiv/2609.00870
curl -X POST /api/v1/records/arxiv/2609.00870/vote -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
  -d '{"value": 1, "emoji": "🤔", "note": "optional"}'
curl -X POST /api/v1/records/arxiv/2609.00870/comments -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
  -d '{"body": "…", "review": true}'