Photonic reservoir computing with dimensionally compressed readout
本文针对硬件限制下的读出层尺寸问题，采用随机投影方法压缩高维水库状态，评估并展示了该方法在特定压缩范围内的优越性能。
0 citationsRead paper
Selected work
Representative Papers
Detecting Historical Turning Points in Italian Media: A Complex Systems Approach to a Diachronic News Corpus
This study addresses the scarcity of annotated historical Italian news corpora by proposing an unsupervised method to automatically identify major socio-political turning points. Building on a diachronic corpus of approximately 600,000 articles from *La Repubblica* (1985–2000), the approach integrates natural language processing, word embeddings, semantic change modeling, complex network analysis, and tools from statistical physics. For the first time, complex systems theory is applied to diachronic media analysis in the Italian context. Without relying on any prior labels, the method successfully detects abrupt shifts in media discourse corresponding to pivotal events such as the transition from Italy’s First to Second Republic, the Gulf War, and the Kosovo War. This work offers a novel paradigm for digital humanities and computational social science by demonstrating how unlabeled textual data can reveal historically significant societal transformations.
0 citationsRead paper
Information-theoretic analysis of temporal dependence in discrete stochastic processes: Application to precipitation predictability
This study quantifies temporal dependencies in the stochastic process of precipitation to enhance forecasting skill and modeling efficiency. We propose a novel information-theoretic metric—“predictability gain”—to characterize higher-order memory effects; integrate block entropy estimation, Bootstrap resampling, and Fisher’s method for p-value combination to achieve robust memory detection under small-sample conditions, overcoming limitations of conventional criteria (e.g., AIC/BIC). Applied to daily precipitation data across the contiguous United States, results confirm predominant low-order Markov behavior, with memory strength exhibiting significant spatiotemporal heterogeneity: strongest in winter along the West Coast and in summer over the Southeast—consistent with underlying climatological dynamics. The framework provides a concise, robust paradigm for stochastic precipitation modeling and real-time predictability assessment.
0 citationsRead paper
Functional Connectivity Networks for Transportation Delay Analysis: from Theory to Software
Existing studies on traffic delay propagation lack a unified functional network modeling framework; node definitions and edge construction often rely on subjective judgment, hindering reproducibility and mechanistic interpretation. This paper proposes a principled functional connectivity network framework for delay propagation: stations serve as nodes, while directed edges are inferred by jointly leveraging dynamic correlation (Pearson) and causality (Granger causality), with explicit clarification of the full modeling pipeline and critical methodological pitfalls. We release *delaynet*, an open-source Python toolkit enabling end-to-end standardized implementation. Empirical evaluation on real-world Swiss railway data demonstrates that the framework accurately reconstructs delay propagation pathways, uncovers hierarchical diffusion structures, and substantially improves analytical reproducibility and interpretability. The approach provides a scalable methodology and software foundation for traffic resilience assessment and proactive control.
0 citationsRead paper
Quantum reservoir computing using Jaynes-Cummings model
This study addresses the limited temporal information processing capability of quantum systems by proposing a quantum reservoir computing framework based on the Jaynes–Cummings model and its dispersive limit. Leveraging high-dimensional Hilbert spaces and intrinsic nonlinear quantum dynamics, the system significantly outperforms classical linear reservoirs on nonlinear memory tasks and chaotic time-series prediction (e.g., Mackey–Glass). Key contributions include: (i) the first demonstration that nonlinear memory capacity can surpass the linear upper bound; (ii) identification of synergistic enhancement between higher-order bosonic operator excitations and temporal multiplexing for representation power; and (iii) experimental confirmation that high-performance temporal modeling is achievable with minimal spin–boson coupling configurations. Rigorous validation confirms superior prediction accuracy, parameter robustness, and physical feasibility, establishing a resource-efficient paradigm for quantum machine learning.
0 citationsRead paper
Recent publications
Latest Papers
Photonic reservoir computing with dimensionally compressed readout
本文针对硬件限制下的读出层尺寸问题，采用随机投影方法压缩高维水库状态，评估并展示了该方法在特定压缩范围内的优越性能。
0 citationsRead paper
Detecting Historical Turning Points in Italian Media: A Complex Systems Approach to a Diachronic News Corpus
This study addresses the scarcity of annotated historical Italian news corpora by proposing an unsupervised method to automatically identify major socio-political turning points. Building on a diachronic corpus of approximately 600,000 articles from *La Repubblica* (1985–2000), the approach integrates natural language processing, word embeddings, semantic change modeling, complex network analysis, and tools from statistical physics. For the first time, complex systems theory is applied to diachronic media analysis in the Italian context. Without relying on any prior labels, the method successfully detects abrupt shifts in media discourse corresponding to pivotal events such as the transition from Italy’s First to Second Republic, the Gulf War, and the Kosovo War. This work offers a novel paradigm for digital humanities and computational social science by demonstrating how unlabeled textual data can reveal historically significant societal transformations.
0 citationsRead paper
Information-theoretic analysis of temporal dependence in discrete stochastic processes: Application to precipitation predictability
This study quantifies temporal dependencies in the stochastic process of precipitation to enhance forecasting skill and modeling efficiency. We propose a novel information-theoretic metric—“predictability gain”—to characterize higher-order memory effects; integrate block entropy estimation, Bootstrap resampling, and Fisher’s method for p-value combination to achieve robust memory detection under small-sample conditions, overcoming limitations of conventional criteria (e.g., AIC/BIC). Applied to daily precipitation data across the contiguous United States, results confirm predominant low-order Markov behavior, with memory strength exhibiting significant spatiotemporal heterogeneity: strongest in winter along the West Coast and in summer over the Southeast—consistent with underlying climatological dynamics. The framework provides a concise, robust paradigm for stochastic precipitation modeling and real-time predictability assessment.
0 citationsRead paper
Functional Connectivity Networks for Transportation Delay Analysis: from Theory to Software
Existing studies on traffic delay propagation lack a unified functional network modeling framework; node definitions and edge construction often rely on subjective judgment, hindering reproducibility and mechanistic interpretation. This paper proposes a principled functional connectivity network framework for delay propagation: stations serve as nodes, while directed edges are inferred by jointly leveraging dynamic correlation (Pearson) and causality (Granger causality), with explicit clarification of the full modeling pipeline and critical methodological pitfalls. We release *delaynet*, an open-source Python toolkit enabling end-to-end standardized implementation. Empirical evaluation on real-world Swiss railway data demonstrates that the framework accurately reconstructs delay propagation pathways, uncovers hierarchical diffusion structures, and substantially improves analytical reproducibility and interpretability. The approach provides a scalable methodology and software foundation for traffic resilience assessment and proactive control.
0 citationsRead paper
Quantum reservoir computing using Jaynes-Cummings model
This study addresses the limited temporal information processing capability of quantum systems by proposing a quantum reservoir computing framework based on the Jaynes–Cummings model and its dispersive limit. Leveraging high-dimensional Hilbert spaces and intrinsic nonlinear quantum dynamics, the system significantly outperforms classical linear reservoirs on nonlinear memory tasks and chaotic time-series prediction (e.g., Mackey–Glass). Key contributions include: (i) the first demonstration that nonlinear memory capacity can surpass the linear upper bound; (ii) identification of synergistic enhancement between higher-order bosonic operator excitations and temporal multiplexing for representation power; and (iii) experimental confirmation that high-performance temporal modeling is achievable with minimal spin–boson coupling configurations. Rigorous validation confirms superior prediction accuracy, parameter robustness, and physical feasibility, establishing a resource-efficient paradigm for quantum machine learning.
0 citationsRead paper