arXiv subjects

Thai Nguyen

Publications and source records attributed to Thai Nguyen.

14 recordsLinked to original sources

arXiv2026

Entity-Constrained CBCT Retrieval for Low-Resource Dental Record Completion

Completing dental records from cone-beam computed tomography (CBCT) is difficult when annotation is scarce and individual clinical fields are supported by different types of evidence. MMDental Task 3 requires seven-field record completion from only 50 labeled CBCT cases and scores the correctness of structured FDI positions and ICD codes; consequently, a visually plausible retrieved record can still be harmful when it introduces an unsupported entity. We propose Entity-Constrained CBCT-Guided Retrieval (ECCR), a parameter-free framework that separates evidence availability from evidence authority. A corpus-derived prior first supplies the complete record. A frozen 3D encoder retrieves image-conditioned Diagnosis evidence, which is appended only if it does not expand the prior FDI or ICD entity set, so the asserted entity set is invariant by construction. On public validation, ECCR reaches a weighted score of 0.3134, improving on both full-record multimodal retrieval (0.2237) and a static text-only prior (0.2915); the guard blocks 63.3% of retrieved candidates, each of which would otherwise have injected an FDI position or ICD code absent from the prior. On the final test evaluation, ECCR obtains 11.37 of a 97.4-point attainable maximum, securing second place overall. The result indicates that, in an extreme low-resource setting, controlling what multimodal evidence is allowed to modify can be more reliable than transferring an entire retrieved record.

cs.CV↗

arXiv2026

Optimal Investment and Entropy-Regularized Learning Under Stochastic Volatility Models with Portfolio Constraints

We study the problem of optimal portfolio selection under stochastic volatility within a continuous time reinforcement learning framework with portfolio constraints. Exploration is modeled through entropy-regularized relaxed controls, where the investor selects probability distributions over admissible portfolio allocations rather than deterministic strategies. Using dynamic programming arguments, we derive the associated entropy-regularized Hamilton-Jacobi-Bellman equation, whose Hamiltonian involves optimization over probability measures supported on a compact control set. We show that the optimal exploratory policy takes the form of a truncated Gaussian distribution characterized by spatial derivatives of the solution of the resulting nonlinear quasilinear parabolic partial differential equation. Under suitable structural conditions on the model coefficients, we prove the existence of classical solutions to this nonlinear HJB equation for the value function. We then establish a verification theorem and analyze the policy-improvement structure induced by the entropy-regularized Hamiltonian, showing how the resulting sequence of PDEs provides a continuous-time interpretation of actor-critic learning dynamics. Finally, our PDE analysis with a semi-closed form of optimal value and optimal policy enables the design of an implementable reinforcement learning algorithm by recasting the optimal problem in a martingale framework.

q-fin.MF↗

arXiv2026

Outperforming a Benchmark with $\alpha$-Bregman Wasserstein divergence

We consider the problem of active portfolio management, where an investor seeks the portfolio with maximal expected utility of the difference between the terminal wealth of their strategy and a proportion of the benchmark's, subject to a budget and a deviation constraint from the benchmark portfolio. As the investor aims at outperforming the benchmark, they choose a divergence that asymmetrically penalises gains and losses as well as penalises underperforming the benchmark more than outperforming it. This is achieved by the recently introduced $\alpha$-Bregman-Wasserstein divergence, subsuming the Bregman-Wasserstein and the popular Wasserstein divergence. We prove existence and uniqueness, characterise the optimal portfolio strategy, and give explicit conditions when the divergence constraints and the budget constraints are binding. We conclude with a numerical illustration of the optimal quantile function in a geometric Brownian motion market model.

q-fin.PM↗

arXiv2026

Pareto and Bowley Reinsurance Games in Peer-to-Peer Insurance

We propose a peer-to-peer (P2P) insurance scheme comprising a risk-sharing pool and a reinsurer. A plan manager determines how risks are allocated among members and ceded to the reinsurer, while the reinsurer sets the reinsurance loading. Our work focuses on the strategic interaction between the plan manager and the reinsurer, and this focus leads to two game-theoretic contract designs: a Pareto design and a Bowley design, for which we derive closed-form optimal contracts. In the Pareto design, cooperation between the reinsurer and the plan manager leads to multiple Pareto-optimal contracts, which are further refined by introducing the notion of coalitional stability. In contrast, the Bowley design yields a unique optimal contract through a leader-follower framework, and we provide a rigorous verification of the individual rationality constraints via pointwise comparisons of payoff vectors. Comparing the two designs, we prove that the Bowley-optimal contract is never Pareto optimal and typically yields lower total welfare. In our numerical examples, the presence of reinsurance improves welfare, especially with Pareto designs and a less risk-averse reinsurer. We further analyze the impact of the single-loading restriction, which disproportionately favors members with riskier losses.

cs.GT↗

arXiv2025

Galton's Law of Mediocrity: Why Large Language Models Regress to the Mean and Fail at Creativity in Advertising

Large language models (LLMs) generate fluent text yet often default to safe, generic phrasing, raising doubts about their ability to handle creativity. We formalize this tendency as a Galton-style regression to the mean in language and evaluate it using a creativity stress test in advertising concepts. When ad ideas were simplified step by step, creative features such as metaphors, emotions, and visual cues disappeared early, while factual content remained, showing that models favor high-probability information. When asked to regenerate from simplified inputs, models produced longer outputs with lexical variety but failed to recover the depth and distinctiveness of the originals. We combined quantitative comparisons with qualitative analysis, which revealed that the regenerated texts often appeared novel but lacked true originality. Providing ad-specific cues such as metaphors, emotional hooks and visual markers improved alignment and stylistic balance, though outputs still relied on familiar tropes. Taken together, the findings show that without targeted guidance, LLMs drift towards mediocrity in creative tasks; structured signals can partially counter this tendency and point towards pathways for developing creativity-sensitive models.

cs.AI↗

arXiv2024

Continuous-time optimal investment with portfolio constraints: a reinforcement learning approach

In a reinforcement learning (RL) framework, we study the exploratory version of the continuous time expected utility (EU) maximization problem with a portfolio constraint that includes widely-used financial regulations such as short-selling constraints and borrowing prohibition. The optimal feedback policy of the exploratory unconstrained classical EU problem is shown to be Gaussian. In the case where the portfolio weight is constrained to a given interval, the corresponding constrained optimal exploratory policy follows a truncated Gaussian distribution. We verify that the closed form optimal solution obtained for logarithmic utility and quadratic utility for both unconstrained and constrained situations converge to the non-exploratory expected utility counterpart when the exploration weight goes to zero. Finally, we establish a policy improvement theorem and devise an implementable reinforcement learning algorithm by casting the optimal problem in a martingale framework. Our numerical examples show that exploration leads to an optimal wealth process that is more dispersedly distributed with heavier tail compared to that of the case without exploration. This effect becomes less significant as the exploration parameter is smaller. Moreover, the numerical implementation also confirms the intuitive understanding that a broader domain of investment opportunities necessitates a higher exploration cost. Notably, when subjected to both short-selling and money borrowing constraints, the exploration cost becomes negligible compared to the unconstrained case.

q-fin.MF↗

arXiv2024

On short-time behavior of implied volatility in a market model with indexes

This paper investigates short-term behaviors of implied volatility of derivatives written on indexes in equity markets when the index processes are constructed by using a ranking procedure. Even in simple market settings where stock prices follow geometric Brownian motion dynamics, the ranking mechanism can produce the observed term structure of at-the-money (ATM) implied volatility skew for equity indexes. Our proposed models showcase the ability to reconcile two seemingly contradictory features found in empirical data from equity markets: the long memory of volatilities and the power law of ATM skews. Furthermore, the models allow for the capture of a new phenomenon termed the quasi-blow-up phenomenon.

q-fin.PR↗

arXiv2022

MLPerf Mobile Inference Benchmark

This paper presents the first industry-standard open-source machine learning (ML) benchmark to allow perfor mance and accuracy evaluation of mobile devices with different AI chips and software stacks. The benchmark draws from the expertise of leading mobile-SoC vendors, ML-framework providers, and model producers. It comprises a suite of models that operate with standard data sets, quality metrics and run rules. We describe the design and implementation of this domain-specific ML benchmark. The current benchmark version comes as a mobile app for different computer vision and natural language processing tasks. The benchmark also supports non-smartphone devices, such as laptops and mobile PCs. Benchmark results from the first two rounds reveal the overwhelming complexity of the underlying mobile ML system stack, emphasizing the need for transparency in mobile ML performance analysis. The results also show that the strides being made all through the ML stack improve performance. Within six months, offline throughput improved by 3x, while latency reduced by as much as 12x. ML is an evolving field with changing use cases, models, data sets and quality targets. MLPerf Mobile will evolve and serve as an open-source community framework to guide research and innovation for mobile AI.

cs.LG↗

arXiv2021

Non-concave expected utility optimization with uncertain time horizon

We consider an expected utility maximization problem where the utility function is not necessarily concave and the time horizon is uncertain. We establish a necessary and sufficient condition for the optimality for general non-concave utility function in a complete financial market. We show that the general concavification approach of the utility function to deal with non-concavity, while being still applicable when the time horizon is a stopping time with respect to the financial market filtration, leads to sub-optimality when the time horizon is independent of the financial risk, and hence can not be directly applied. For the latter case, we suggest a recursive procedure which is based on the dynamic programming principle. We illustrate our findings by carrying out a multi-period numerical analysis for optimal investment problem under a convex option compensation scheme with random time horizon. We observe that the distribution of the non-concave portfolio in both certain and uncertain random time horizon is right-skewed with a long right tail, indicating that the investor expects frequent small losses and a few large gains from the investment. While the (certain) average time horizon portfolio at a premature stopping date is unimodal, the random time horizon portfolio is multimodal distributed which provides the investor a certain flexibility of switching between the local maximizers, depending on the market performance. The multimodal structure with multiple peaks of different heights can be explained by the concavification procedure, whereas the distribution of the time horizon has significant impact on the amplitude between the modes.

q-fin.PM↗

arXiv2020

Utility maximization under endogenous pricing

We study the expected utility maximization problem of a large investor who is allowed to make transactions on tradable assets in an incomplete financial market with endogenous permanent market impacts. The asset prices are assumed to follow a nonlinear price curve quoted in the market as the utility indifference curve of a representative liquidity supplier. Using generalized subgradients, we show that optimality can be fully characterized via a system of coupled forward-backward stochastic differential equations (FBSDEs) which corresponds to a non-linear backward stochastic partial differential equation (BSPDE). We show existence of solutions to the optimal investment problem and the FBSDEs in the case where the driver function of the representative market maker grows at least quadratically or the utility function of the large investor falls faster than quadratically or is exponential. Furthermore, we derive smoothness results for the existence of solutions of BSPDEs. Examples are provided when the market is complete, the driver is positively homogeneous or the utility function is exponential.

q-fin.MF↗

arXiv2019

Optimal investment for participating insurance contracts under VaR-Regulation

This paper studies a Value-at-Risk (VaR)-regulated optimal portfolio problem of the equity holders of a participating life insurance contract. In a setting with unhedgeable mortality risk and complete financial market, the optimal solution is given explicitly for contracts with mortality risk using a martingale approach for constrained non-concave optimization problems. We show that regulatory VaR constraints for participating insurance contracts lead to more prudent investment than in the case of no regulation. This result is contrary to the situation where the insurer maximizes the utility of the total wealth of the company (without distinguishing between contributions of equity holders and policyholders), in which case a VaR constraint may induce the insurer to take excessive risks leading to higher losses than in the case of no regulation. Compared to the unregulated problem, the VaR-constrained strategy leads to a higher expected utility for the policyholders, highlighting the potential usefulness of a VaR-regulation in the context of insurance. The prudent investment behavior is more significant if a VaR-type regulation is replaced by a portfolio insurance (PI)-type regulation. Furthermore, a stricter regulation (a smaller allowed default probability in the VaR problem or a higher minimum guarantee level in the PI problem) enhances the benefit of the policyholder but deteriorates that of the insurer. For both types of regulation, the gains in terms of expected utility are greater for higher participation rates, while being smaller for higher bonus rates. We also extend our analysis to frameworks where dividend and premature death benefit payments are made at an intermediate time date.

q-fin.MF↗

arXiv2018

Plastic flow and localization in an amorphous material: experimental interpretation of the fluidity

We present a thorough study of the plastic response of a granular material progressively loaded. We study experimentally the evolution of the plastic field from a homogeneous one to an heterogeneous one and its fluctuations in term of incremental strain. We show that the plastic field can be decomposed in two components evolving on two decoupled strain increment scales. We argue that the slowly varying part of the field can be identified to the so-called fluidity field introduced recently to interpret the rheological behaviour of amorphous materials. This fluidity field progressively concentrates along a macroscopic direction corresponding to the Mohr-Coulomb angle.

cond-mat.soft↗

arXiv2016

Optimal investment and consumption with downside risk constraint in jump-diffusion models

This paper extends the results of the article [C. Klüppelberg and S. M. Pergamenchtchikov. Optimal consumption and investment with bounded downside risk for power utility functions. In Optimality and Risk: {\it Modern Trends in Mathematical Finance. The Kabanov Festschrift}, pages 133-169, 2009] to a jump-diffusion setting. We show that under the assumption that only positive jumps in the asset prices are allowed, the explicit optimal strategy can be found in the subset of admissible strategies satisfying the same risk constraint as in the pure diffusion setting. When negative jumps probably happen, the regulator should be more conservative. In that case, we suggest to impose on the investor's portfolio a stricter constraint which depends on the probability of having negative jumps in the assets during the whole considered horizon.

q-fin.MF↗

arXiv2014

A Computational Model of Crowds for Collective Intelligence

In this work, we present a high-level computational model of IT-mediated crowds for collective intelligence. We introduce the Crowd Capital perspective as an organizational-level model of collective intelligence generation from IT-mediated crowds, and specify a computational system including agents, forms of IT, and organizational knowledge.

cs.CY↗