Structural Predictors of Power-Domination Propagation Dynamics in Complex Networks: An Explainable Machine-Learning Approach

Power domination models network monitoring through an initial domination step followed by iterative propagation. Although previous studies have mainly focused on identifying small or minimum power-dominating sets, candidate sets of similar size can exhibit substantially different propagation behavior. The structural factors associated with these differences remain less well understood. This study investigates the structural predictors of power-domination success and power propagation time across Erdős–Rényi, Barabási–Albert, and Watts–Strogatz networks using explainable machine learning. Controlled simulations are conducted across multiple network sizes, candidate-set fractions, and selection strategies. Two complementary tasks are considered: classification of PDS success and regression of propagation time among successful configurations. The analysis combines graph-level train–test partitioning, SHAP-based interpretation, feature ablation, repeated validation, cross-topology transfer, and distribution-shift analysis. The results show that structural models outperform compact baseline models, particularly for propagation-time prediction. Initial monitoring coverage provided the strongest predictive signal, although part of this predictive relationship follows directly from the definition of the propagation process, and average path length emerged as the most prominent additional structural predictor in the primary analysis. Candidate-set configuration was also associated with both monitoring success and propagation speed. Cross-topology transfer showed stronger performance on unseen ER and BA networks than on unseen WS networks. A symmetric graph-level distribution-shift analysis showed a covariate shift in all three transfer directions, with the weaker transfer to WS coinciding with broader full-range extrapolation across several measured structural predictors. Overall, power-domination dynamics are jointly associated with initial coverage, global network structure, and candidate-set configuration, providing an interpretable perspective beyond minimum-set optimization.

Authors

Institutions

Publication Details

Journal
Electronics
Published
2026-11-20
DOI
https://doi.org/10.3390/electronics15225418
Primary Topic
Complex Network Analysis Techniques
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Structural Predictors of Power-Domination Propagation Dynamics in Complex Networks: An Explainable Machine-Learning Approach

Bader Alwasel
Electronics
Complex Network Analysis Techniques
article

Structural Predictors of Power-Domination Propagation Dynamics in Complex Networks: An Explainable Machine-Learning Approach

Bader Alwasel
article en

Abstract

Power domination models network monitoring through an initial domination step followed by iterative propagation. Although previous studies have mainly focused on identifying small or minimum power-dominating sets, candidate sets of similar size can exhibit substantially different propagation behavior. The structural factors associated with these differences remain less well understood. This study investigates the structural predictors of power-domination success and power propagation time across Erdős–Rényi, Barabási–Albert, and Watts–Strogatz networks using explainable machine learning. Controlled simulations are conducted across multiple network sizes, candidate-set fractions, and selection strategies. Two complementary tasks are considered: classification of PDS success and regression of propagation time among successful configurations. The analysis combines graph-level train–test partitioning, SHAP-based interpretation, feature ablation, repeated validation, cross-topology transfer, and distribution-shift analysis. The results show that structural models outperform compact baseline models, particularly for propagation-time prediction. Initial monitoring coverage provided the strongest predictive signal, although part of this predictive relationship follows directly from the definition of the propagation process, and average path length emerged as the most prominent additional structural predictor in the primary analysis. Candidate-set configuration was also associated with both monitoring success and propagation speed. Cross-topology transfer showed stronger performance on unseen ER and BA networks than on unseen WS networks. A symmetric graph-level distribution-shift analysis showed a covariate shift in all three transfer directions, with the weaker transfer to WS coinciding with broader full-range extrapolation across several measured structural predictors. Overall, power-domination dynamics are jointly associated with initial coverage, global network structure, and candidate-set configuration, providing an interpretable perspective beyond minimum-set optimization.

ElectronicsVol. 15(22)
Qassim University (SA)
Qassim University
Complex Network Analysis Techniques
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon
Structural Predictors of Power-Domination Propagation Dynamics in Complex Networks: An Explainable Machine-Learning Approach — Bader Alwasel · Electronics (2026) | TGRS Research Map | TGRS