AI Paper Atlas

Research map / Learning theory and algorithms / Prediction and decision losses

Insights into Ordinal Embedding Algorithms: A Systematic Evaluation

Leena Chennuru Vankadara, Michael Lohaus, Siavash Haghiri, Faiz Ul Wahab, Ulrike von Luxburg

JMLR 2023 · Learning theory and algorithms › Prediction and decision losses · approach: Metric and ordinal embedding

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use t-STE instead of STESTE · MNISTtest triplet error: 0.11 -> 0.2Hurt
use t-STE instead of STESTE · COVTYPEkNN error: 0.5 -> 0.53Hurt
optimize over the embedding matrix (SOE) instead of Gram matrix (GNMDS)GNMDS · USPStest triplet error: 0.35 -> 0.02Improved
optimize over the embedding matrix (SOE) instead of Gram matrix (CKL)CKL · USPSkNN error: 0.45 -> 0.04Improved
use FORTE instead of other Gram-matrix methodsGNMDS · USPStest triplet error: 0.35 -> 0.06Improved
use SOE instead of landmark method LLOELLOE · MNISTtest triplet error: 0.31 -> 0.03Improved
use SOE instead of neural method OENNOENN · MNISTtest triplet error: 0.13 -> 0.03Improved
use SOE instead of tSTEtSTE · COVTYPEtest triplet error: 0.24 -> 0.02Improved
use SOE instead of OENN on COVTYPEOENN · COVTYPEkNN error: 0.29 vs 0.13Hurt
choose triplets at random instead of the landmark selection procedure for SOESOE · generalconvergence timeImproved
increase noise level in tripletsall OE methods · Aggregationreconstruction qualityHurt
increase embedding dimensionSOE, STE, tSTE, OENN · USPSkNN and Procrustes errorImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy