5 papers
Learning Source Acquisition Policies by Offline Planning
Ziqi Zhao, Run Xu, Qingjian Ni
Predicting under an acquisition budget requires choosing feature groups whose value can depend on later queries. O-MPAC transfers finite-horizon risk-cost targets from complete tra…
Matched Queries for Curvature and Density at Branching Junctions
Ziqi Zhao, Qingjian Ni
At a junction, a score field can reveal weighted tangent rays, yet these first-order quantities do not determine how individual branches bend or how their densities change away fro…
Recovering Weighted Tangent Geometry from a Single-Scale Score Field
Ziqi Zhao, Qingjian Ni
Near a smooth data manifold, one tangent space summarizes local geometry. At a branch point, the corresponding first-order object is instead a measure over tangent directions, whos…
When More Generators Hurt: Shellsort on Full Product Grids
Ziqi Zhao, Qingjian Ni
Shellsort repeatedly runs insertion sort with decreasing gaps, so its worst-case cost depends on the gap sequence. Pratt's sequence, one of the few systematic construction…
A Multi-scale Representation Learning Framework for Long-Term Time Series Forecasting
Boshi Gao, Qingjian Ni, Fanbo Ju +2
Long-term time series forecasting (LTSF) offers broad utility in practical settings like energy consumption and weather prediction. Accurately predicting long-term changes, however…