.png){kind=icon}
.png){kind=icon}
Return
A scalable descent algorithm for network design problems
.png){kind=icon}
PRE
AI
DOI：10.1016/j.tre.2026.104838.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• We propose a scalable descent algorithm for solving network design problems. • We show that the algorithm is more scalable and efficient than all existing methods that can compute the gradient of such problem exactly. • We demonstrate that the algorithm requires memory comparable to a state-of-theart heuristic that only approximates gradients. • We conduct large-scale experiments on realistic networks to demonstrate the practical applicability of the approach. • We show that accurate gradient computation is especially critical in highly congested networks.
Keywords:
scalable descent algorithm
network design problems
gradient computation
large-scale experiments
congested networks
Journal
T
IF:
0
Papers:
247
Citations:
0