Multi-location networks and ecosystem detection: methods and applications

Sabharwal, Siddhanth

Permalink

https://hdl.handle.net/2142/130140 Copy

Description

Title

Multi-location networks and ecosystem detection: methods and applications

Author(s)

Sabharwal, Siddhanth

Issue Date

2025-07-07

Director of Research (if dissertation) or Advisor (if thesis)

Chen, Yuguo

Doctoral Committee Chair(s)

Chen, Yuguo

Committee Member(s)

• Culpepper, Steven A
• Vargas, Pamela M
• Zhang, Susu

Department of Study

Statistics

Discipline

Statistics

Degree Granting Institution

University of Illinois Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• network
• networks
• graph
• graphs
• multi-location
• multilocation
• multi-layer
• multilayer
• multiplex
• ecosystem
• ecosystems

Abstract

We introduce multi-location networks to model the real-world phenomena of nodes existing in multiple places simultaneously. A multi-location network contains a set of locations. Each location can have a different set of nodes, but nodes can be present in multiple locations at once. In contrast, a multiplex network contains a set of layers, where each layer shares the same set of nodes. By combining these two structures, a multi-location multiplex network (MLMN) enables nodes to appear in various locations and connect through multiple edge types. For example, in a species interaction MLMN, locations could correspond to continents, while layers represent different types of interactions, such as predation and competition. Some species may inhabit multiple continents, with interactions varying by location. Analyzing the evolution of MLMNs poses substantial challenges due to the complexity of integrating location and edge type. Traditional dynamic network logistic regression models often aggregate data across these structures, leading to information loss. In this dissertation, we extend dynamic network logistic regression to accommodate MLMNs, improving the estimation of factors driving changes in such networks. Our approach predicts node presence in each location and the formation of edges across different layers at each location. We provide theoretical guarantees for the model, including global optimality, consistency, asymptotic normality, and asymptotic efficiency of the estimated parameters. The model's finite sample performance is evaluated through simulations. Additionally, a case study is conducted using a dataset from Equinix, a 2024 Fortune 500 company and one of the world's largest data center operators. Our findings highlight the strong impact of location on node presence and link formation, showcasing the practical relevance of understanding multi-location structures to capture complex network dynamics. Dynamic Generalized Multi-Location Multiplex Networks (DGMLMNs) describe a sophisticated class of networks characterized by temporal node dynamics, directed weighted edges, and concurrent multiple networks indexed by location, with nodes allowed to be in multiple locations at once. This research addresses a gap in network modeling literature by developing methods to describe these complex network structures. We propose a comprehensive approach to infer key drivers of node presence and link formation, enabling prediction of the future state of a network. The methodology employs a two-step network structure model: first modeling node presence across locations, second modeling edge counts from one node to another for each location-layer combination. Through theoretical analysis, we establish the asymptotic properties of the estimators and through simulations, we validate the finite-sample estimation and prediction performance. The model's practical utility is demonstrated through application on an anonymized DGMLMN dataset from Equinix, a 2024 Fortune 500 company, where the findings allowed for the testing and validation of various business hypotheses. As the development of methods for complex networks continues to expand, our contribution aims to open up new applications of network science for practitioners. Ecosystems, like communities, are groupings of nodes within networks. However, unlike communities, ecosystems are defined not by dense connectivity but by the nature of competition and cooperation. Drawing on ideas from management science literature, we introduce a rigorous network-based framework for defining key ecosystem concepts such as focal nodes, suppliers, customers, and complementors. These concepts give further meaning to terms like value, co-opetition, and ecosystem formation. We propose a new method, ecosystem detection, which takes a directed, weighted network as input and outputs an interpretable partition of the nodes. The method proceeds in three steps: identifying focal nodes, clustering them into ecosystems, and assigning the remaining nodes to an ecosystem. A hypothesis testing procedure to verify the presence of ecosystems within a network is also presented. We apply this method to two real-world networks. The first is the United States (U.S.) airport network, which records the count of passenger flights between airports. The second is an anonymized interconnection network provided by Equinix, a global data center operator and 2024 Fortune 500 company. In both cases, ecosystem detection yields contextually coherent partitions, offering insights beyond those provided by community detection.

Graduation Semester

2025-08

Type of Resource

Thesis

Handle URL

https://hdl.handle.net/2142/130140

Copyright and License Information

Copyright 2025 Siddhanth Sabharwal

Owning Collections