Sparse Network Inference under Imperfect Detection and its Application to Ecological Networks
2026-04-22T07:24:06Z•99b95fd12c39b6570238b0071e13f950fb19739dafe09a907205fc5489deff76
ADMMGaussian-mixture-modelsLLaDA-fine-tuning','gradient-boosting','GAM','ParamBoost','modelSobol-indicesarxivcontinuous-time-RLcuriositydiscrete-tilt-matchingecological-networksfine-tuninggenerator-regressionintrinsic-rewardsl1/2-regularizationlikelihood-freemachine-learningmasked-diffusion-LLMsmodel-selectionnonconvex-optimizationpolicy-evaluationpolynomial-chaos-expansionsingular-value-thresholdingsparse-network-inferencestatisticsuncertainty-quantificationworld-models
What happened
An RSS batch of new arXiv submissions (stat.ML / cross-listed) covering a range of methodological advances in machine learning and statistics. Key topics include: sparse network inference under imperfect detection using nonconvex ℓ1/2 regularization and an ADMM solver; high-order generator regression for continuous-time policy evaluation; a fast SVD-based estimator for the number of Gaussian mixture components; analytical extraction of conditional Sobol' indices from Polynomial Chaos Expansions; a Curiosity-Critic intrinsic reward for world-model training; Discrete Tilt Matching (DTM) for fine
Why it matters
A reviewed impact interpretation has not been published for this record.
Evidence and limitations
- Source ID
- arxiv_stat_ml
- Record identifier
- 99b95fd12c39b6570238b0071e13f950fb19739dafe09a907205fc5489deff76
- Enrichment time
- 2026-04-22T07:24:06Z
- AI-assisted enrichment
- Yes
This record may overlap with other records. Its enrichment can be incomplete or wrong, and machine assistance was used. Validate consequential decisions against the linked source and your own environment.