On-device research index

arXiv research

A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.

168,742 papers · 148 categories

Trend · papers per month

count % of corpus

2 results for “scale-insensitive”

Scale-Insensitive Neural Network Significance Testsstat.ML

Develops a new method for neural network significance testing without strict constraints.

problem Testing neural networks without bounded weights or specific architectural constraints.

method Uses Rademacher complexity bounds, weakened Sobolev space membership conditions, and a modified sieve space construction.

result Achieves optimal convergence rates and valid asymptotic distributions for test statistics.

neural networks significance testing scale-insensitive Rademacher complexity Sobolev space

2025-01-27abs ↗pdf ↗

Towards Better Forecasting by Fusing Near and Distant Future Visionscs.LG

Proposes MLCNN for better multivariate time series forecasting.

problem Challenges in forecasting multivariate time series, especially the limitation of predicting only one future moment.

method MLCNN, a multi-task deep learning framework inspired by Construal Level Theory, fuses future visions of near and distant future predictions.

result Significant improvements in forecasting accuracy (4.59% RMSE reduction, 6.87% MAE reduction) on real-world datasets.

time series forecasting multivariate forecasting future visions neural networks psychology

2019-12-11abs ↗pdf ↗