A delay-Adaptive neural network for querying TOP-k critical vertices on time-Dependent shortest paths

Area of Science:

• Computer Science
• Network Science
• Artificial Intelligence

Background:

• Real-world networks exhibit temporal dependencies, making static models insufficient for critical vertex queries.
• Existing methods for top-k critical vertices (kCV) queries are often limited to static networks.
• Time-dependent shortest paths require advanced algorithms to account for changing network conditions.

Purpose of the Study:

• To propose a delay-adaptive neural network (DANN) for querying top-k critical vertices (kCV) on time-dependent shortest paths.
• To formally define the mathematical model for kCV queries on time-dependent networks.
• To design and implement delay-adaptive neurons and DANN operational mechanisms.

Main Methods:

• Development of a weightless computational network (DANN) with a physical network topology mapping structure.
• Categorization of delay-adaptive neurons into source-peripheral, destination-peripheral, and intermediate types.
• Implementation of deterministic logic gates within neurons for information processing and parallel computing capabilities.

Main Results:

• The DANN requires no training and demonstrates strong adaptability across networks of varying scales and structures.
• The DANN architecture ensures parallel computing and synchronization, enhancing query speed and guaranteeing optimal solutions.
• Comparative experiments on real road network data validate the effectiveness and advanced nature of the proposed approach.

Conclusions:

• The proposed DANN provides an effective and efficient solution for kCV queries on time-dependent networks.
• The weightless and adaptive nature of DANN makes it suitable for dynamic and large-scale network analysis.
• This research advances the field of network analysis by offering a novel approach to critical vertex identification in temporal networks.