A learned data structure maintains (1+epsilon)-approximate single-source shortest paths under edge insertions in O~(m min_tau {tau+|HIGH(tau)|} log W/epsilon) total time, with an offline building block running in near-linear time.
Bercea, Jakob Bæk Tejs Houen, and Rasmus Pagh
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.DS 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Incremental Approximate Single-Source Shortest Paths with Predictions
A learned data structure maintains (1+epsilon)-approximate single-source shortest paths under edge insertions in O~(m min_tau {tau+|HIGH(tau)|} log W/epsilon) total time, with an offline building block running in near-linear time.