Multi-layer network feature fusion XGBoost model author collaboration prediction data

This dataset serves as the supporting data for the paper "Research on Author Collaboration Relationship Prediction Based on Multi-layer Network Feature Fusion XGBoost Model" (manuscript number: 2026-0106, "Data Analysis and Knowledge Discovery"), and includes four files:​(1) Multilayer Network_1000 Rows Sampling: This paper constructs a four-layer network of "Country - Institution - Author - Topic" based on medical informatics literature data from 2004 to 2023, and the file is a sample of the network layer data of the four-layer network.​(2) Distribution Histogram and Log-Log Plot of the Number of Collaborations among Author Pairs (Dependent Variable): To clarify the overall characteristics of author collaborations in the field of medical informatics, this paper conducts descriptive statistical analysis on the distribution of the number of collaborations among author pairs, the core dependent variable.

The file contains the raw data on the distribution of the number of collaborations among authors in this field.​(3) Author Collaboration Prediction Model: The file contains hyperparameters of the XGBoost model and other comparison models, prediction result data for author collaboration relationships, prediction result data for author collaboration intensity, including accuracy, precision, recall, and F1 score.​(4) Bibliography List for Literature Review: This document is a list of bibliographic entries for the literature involved in the literature review and analysis sections of this paper.