Selected work

Representative Papers

An Artificial Market for Brazilian Real Estate Investment Funds: An Agent-Based Proposal

Jul 27, 2026

This study addresses a critical gap in the literature by developing a computational model capable of capturing the full value chain of Brazilian Real Estate Investment Trusts (FIIs) and their interaction with macroeconomic dynamics, which has previously hindered effective policy analysis and mechanism design. The authors propose an agent-based artificial market that, for the first time, integrates within a unified framework the entire FII process—from property income generation and dividend distribution to trading of shares by heterogeneous investors via a double-auction order book. The model endogenously incorporates key macroeconomic variables such as the Selic interest rate and inflation, alongside behaviorally heterogeneous agents whose decisions are driven by financial literacy. It successfully replicates key stylized facts of the IFIX index, achieving over 75% coverage of calibrated moments and producing simulated trajectories statistically indistinguishable from empirical data in 96% of cases, while spontaneously generating power-law decay in absolute return autocorrelations and aggregate Gaussianity.

0 citationsRead paper

Certified Lower Bounds and Efficient Estimation of Minimum Accuracy in Quantum Kernel Methods

Dec 23, 2025

Existing minimum-accuracy heuristics for quantum kernel methods suffer from high computational cost, applicability only to balanced datasets, and lack of theoretical guarantees. Method: This work generalizes the notion of minimum accuracy to arbitrary binary classification datasets under approximate quantum hardware, rigorously proving it as a theoretical lower bound on the empirical accuracy of linear classifiers. We propose a Pauli-direction-based Monte Carlo estimation technique, providing probabilistic error bounds and formal convergence guarantees. Crucially, our method evaluates quantum feature map quality without training a quantum support vector machine (QSVM). Contribution/Results: The approach significantly reduces computational complexity while ensuring scalability, theoretical soundness, and hardware compatibility. It constitutes the first provably reliable, practical tool for pre-screening quantum feature maps—enabling efficient, theoretically grounded selection of promising quantum embeddings prior to full quantum kernel training.

0 citationsRead paper

$k$-path graphs: experiments and conjectures about algebraic connectivity and $α$-index

Nov 26, 2025

This study investigates two fundamental spectral invariants of $k$-path graphs—the algebraic connectivity (i.e., the second smallest Laplacian eigenvalue) and the $alpha$-spectral radius (i.e., the largest eigenvalue of the $A_alpha$ matrix)—with the aim of characterizing their extremal structures. Method: Building upon and refining Pereira et al.’s generation algorithm, we systematically construct and enumerate all non-isomorphic $k$-path graphs for $k=2,3,4$ across various orders, then perform exhaustive high-precision numerical spectral computations. Contribution/Results: Based on spectral analysis of thousands of graphs, we obtain the complete lists of extremal graphs for both invariants at small orders—first such results in the literature—and formulate several provable conjectures regarding extremal structure, including vertex distribution patterns, branching configurations, and dependence on $alpha$. This work establishes the first large-scale empirical foundation and structural insight into extremal spectral problems for path-based graph families.

0 citationsRead paper

On the distribution of $A_α$-eigenvalues in terms of graph invariants

Oct 08, 2025

This study characterizes the distribution of eigenvalues of the $A_alpha$-matrix of a graph $G$ over real subintervals, aiming to establish quantitative bounds—both upper and lower—on the number of such eigenvalues in terms of structural graph parameters, including the number of pendant vertices, quasi-pendant vertices, domination number, matching number, and edge covering number. Employing algebraic graph theory techniques—including analysis of the characteristic polynomial, Rayleigh quotient estimation, and structural induction—we systematically relate the $A_alpha$-spectrum distribution to multiple graph invariants for the first time. This unifies and generalizes classical spectral bounds for both the adjacency matrix and the signless Laplacian matrix. The derived bounds are tight, and extremal graph families achieving these bounds are explicitly constructed, thereby confirming the optimality and broad applicability of the theoretical results.

0 citationsRead paper

A Machine Learning Approach to Automatic Fall Detection of Combat Soldiers

Jan 26, 2025

To address the real-time detection of incapacitating injuries caused by sudden soldier falls in battlefield environments, this paper proposes a lightweight fall detection method tailored for combat casualty response. We introduce the first publicly available inertial fall dataset specifically designed for military scenarios and develop a one-dimensional convolutional neural network (1D-CNN) discriminative model that fuses wrist- and center-of-mass-mounted inertial sensor data. The model incorporates Bayesian hyperparameter optimization and explicitly models pre-shock physiological indicators associated with traumatic shock. Evaluated under realistic simulated combat conditions, the method achieves 98.7% detection accuracy, a false alarm rate below 0.5%, and end-to-end latency under 300 ms, while demonstrating high robustness and low power consumption. This work delivers the first operationally validated fall recognition solution for wearable battlefield medical response systems.

0 citationsRead paper

Recent publications

Latest Papers

An Artificial Market for Brazilian Real Estate Investment Funds: An Agent-Based Proposal

Jul 27, 2026

This study addresses a critical gap in the literature by developing a computational model capable of capturing the full value chain of Brazilian Real Estate Investment Trusts (FIIs) and their interaction with macroeconomic dynamics, which has previously hindered effective policy analysis and mechanism design. The authors propose an agent-based artificial market that, for the first time, integrates within a unified framework the entire FII process—from property income generation and dividend distribution to trading of shares by heterogeneous investors via a double-auction order book. The model endogenously incorporates key macroeconomic variables such as the Selic interest rate and inflation, alongside behaviorally heterogeneous agents whose decisions are driven by financial literacy. It successfully replicates key stylized facts of the IFIX index, achieving over 75% coverage of calibrated moments and producing simulated trajectories statistically indistinguishable from empirical data in 96% of cases, while spontaneously generating power-law decay in absolute return autocorrelations and aggregate Gaussianity.

0 citationsRead paper

Certified Lower Bounds and Efficient Estimation of Minimum Accuracy in Quantum Kernel Methods

Dec 23, 2025

Existing minimum-accuracy heuristics for quantum kernel methods suffer from high computational cost, applicability only to balanced datasets, and lack of theoretical guarantees. Method: This work generalizes the notion of minimum accuracy to arbitrary binary classification datasets under approximate quantum hardware, rigorously proving it as a theoretical lower bound on the empirical accuracy of linear classifiers. We propose a Pauli-direction-based Monte Carlo estimation technique, providing probabilistic error bounds and formal convergence guarantees. Crucially, our method evaluates quantum feature map quality without training a quantum support vector machine (QSVM). Contribution/Results: The approach significantly reduces computational complexity while ensuring scalability, theoretical soundness, and hardware compatibility. It constitutes the first provably reliable, practical tool for pre-screening quantum feature maps—enabling efficient, theoretically grounded selection of promising quantum embeddings prior to full quantum kernel training.

0 citationsRead paper

$k$-path graphs: experiments and conjectures about algebraic connectivity and $α$-index

Nov 26, 2025

This study investigates two fundamental spectral invariants of $k$-path graphs—the algebraic connectivity (i.e., the second smallest Laplacian eigenvalue) and the $alpha$-spectral radius (i.e., the largest eigenvalue of the $A_alpha$ matrix)—with the aim of characterizing their extremal structures. Method: Building upon and refining Pereira et al.’s generation algorithm, we systematically construct and enumerate all non-isomorphic $k$-path graphs for $k=2,3,4$ across various orders, then perform exhaustive high-precision numerical spectral computations. Contribution/Results: Based on spectral analysis of thousands of graphs, we obtain the complete lists of extremal graphs for both invariants at small orders—first such results in the literature—and formulate several provable conjectures regarding extremal structure, including vertex distribution patterns, branching configurations, and dependence on $alpha$. This work establishes the first large-scale empirical foundation and structural insight into extremal spectral problems for path-based graph families.

0 citationsRead paper

On the distribution of $A_α$-eigenvalues in terms of graph invariants

Oct 08, 2025

This study characterizes the distribution of eigenvalues of the $A_alpha$-matrix of a graph $G$ over real subintervals, aiming to establish quantitative bounds—both upper and lower—on the number of such eigenvalues in terms of structural graph parameters, including the number of pendant vertices, quasi-pendant vertices, domination number, matching number, and edge covering number. Employing algebraic graph theory techniques—including analysis of the characteristic polynomial, Rayleigh quotient estimation, and structural induction—we systematically relate the $A_alpha$-spectrum distribution to multiple graph invariants for the first time. This unifies and generalizes classical spectral bounds for both the adjacency matrix and the signless Laplacian matrix. The derived bounds are tight, and extremal graph families achieving these bounds are explicitly constructed, thereby confirming the optimality and broad applicability of the theoretical results.

0 citationsRead paper

A Machine Learning Approach to Automatic Fall Detection of Combat Soldiers

Jan 26, 2025

To address the real-time detection of incapacitating injuries caused by sudden soldier falls in battlefield environments, this paper proposes a lightweight fall detection method tailored for combat casualty response. We introduce the first publicly available inertial fall dataset specifically designed for military scenarios and develop a one-dimensional convolutional neural network (1D-CNN) discriminative model that fuses wrist- and center-of-mass-mounted inertial sensor data. The model incorporates Bayesian hyperparameter optimization and explicitly models pre-shock physiological indicators associated with traumatic shock. Evaluated under realistic simulated combat conditions, the method achieves 98.7% detection accuracy, a false alarm rate below 0.5%, and end-to-end latency under 300 ms, while demonstrating high robustness and low power consumption. This work delivers the first operationally validated fall recognition solution for wearable battlefield medical response systems.

0 citationsRead paper