Selected work

Representative Papers

Concentration Within Distribution: Unmasking Bitcoin's Structural Centralization Through Network Science

Nov 29, 2025

Although the Bitcoin protocol is designed to be decentralized, its real-world user network (BUN) exhibits a pronounced core–periphery mesoscopic structure, revealing latent structural centralization risks. Method: Leveraging raw blockchain data, we construct a dedicated database and propose directed-sensitive PageRank and HITS centrality measures; we further design four variants of Newman’s assortativity coefficient to enable multidimensional, dynamic characterization of structural influence distribution. Combining connected component analysis with high-frequency price volatility correlation testing, we examine temporal evolution of network topology. Contribution/Results: We identify persistent consolidation among a few dominant connected components, indicating an emergent “concentration-in-distribution” evolutionary trend. Crucially, structural concentration—quantified via our metrics—exhibits statistically significant correlation with market price volatility, providing empirical evidence that topological centralization in the BUN may amplify financial instability. This work delivers the first comprehensive, data-driven analysis of structural centralization dynamics in Bitcoin’s actual user network.

0 citationsRead paper

Determination of galaxy photometric redshifts using Conditional Generative Adversarial Networks (CGANs)

Jan 11, 2025

To address the limited accuracy and poorly quantified uncertainties in photometric redshift (photo-z) estimation for wide-field imaging surveys, this work introduces conditional generative adversarial networks (CGANs) to photo-z estimation for the first time, shifting the paradigm from deterministic point estimates to full posterior probability density function (PDF) modeling. The proposed method takes multi-band photometric measurements as input and leverages CGANs to directly generate well-calibrated redshift posterior distributions under explicit conditioning. Experiments on the Dark Energy Survey (DES) Year 1 dataset demonstrate that our approach significantly improves PDF calibration compared to conventional methods such as random forests. It further exhibits superior robustness to outlier objects and more accurate uncertainty quantification. By delivering statistically well-calibrated, probabilistic distance inference, the framework provides a reliable foundation for large-scale cosmological surveys.

0 citationsRead paper

Recent publications

Latest Papers

Concentration Within Distribution: Unmasking Bitcoin's Structural Centralization Through Network Science

Nov 29, 2025

Although the Bitcoin protocol is designed to be decentralized, its real-world user network (BUN) exhibits a pronounced core–periphery mesoscopic structure, revealing latent structural centralization risks. Method: Leveraging raw blockchain data, we construct a dedicated database and propose directed-sensitive PageRank and HITS centrality measures; we further design four variants of Newman’s assortativity coefficient to enable multidimensional, dynamic characterization of structural influence distribution. Combining connected component analysis with high-frequency price volatility correlation testing, we examine temporal evolution of network topology. Contribution/Results: We identify persistent consolidation among a few dominant connected components, indicating an emergent “concentration-in-distribution” evolutionary trend. Crucially, structural concentration—quantified via our metrics—exhibits statistically significant correlation with market price volatility, providing empirical evidence that topological centralization in the BUN may amplify financial instability. This work delivers the first comprehensive, data-driven analysis of structural centralization dynamics in Bitcoin’s actual user network.

0 citationsRead paper

Determination of galaxy photometric redshifts using Conditional Generative Adversarial Networks (CGANs)

Jan 11, 2025

To address the limited accuracy and poorly quantified uncertainties in photometric redshift (photo-z) estimation for wide-field imaging surveys, this work introduces conditional generative adversarial networks (CGANs) to photo-z estimation for the first time, shifting the paradigm from deterministic point estimates to full posterior probability density function (PDF) modeling. The proposed method takes multi-band photometric measurements as input and leverages CGANs to directly generate well-calibrated redshift posterior distributions under explicit conditioning. Experiments on the Dark Energy Survey (DES) Year 1 dataset demonstrate that our approach significantly improves PDF calibration compared to conventional methods such as random forests. It further exhibits superior robustness to outlier objects and more accurate uncertainty quantification. By delivering statistically well-calibrated, probabilistic distance inference, the framework provides a reliable foundation for large-scale cosmological surveys.

0 citationsRead paper