MTMinh-Ngoc Tran

Minh-Ngoc Tran

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Minh-Ngoc Tran has a total of 64 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Neural-Network Volatility Forecasting

    by Chen Liu, Minh-Ngoc Tran, Chao Wang, Richard Gerlach, Robert Kohn

    We investigate how training regimes and sample size affect the performance of neural networks in financial forecasting. Using volatility forecasts for more than 10,000 stocks, we find that, within the specifications studied, performance gains from sample size scaling outweigh those from model size scaling or architectural choice. We further show that, when NNs are applied to financial data, global estimation is better aligned with their data-intensive nature and materially enhances their practical viability.

  • Open Access

    Economically Disciplined Operator Learning for Stable Commodity Demand Forecasting

    by Hongwei Ma, Ye Xiao, Jiayu Fang, S. T. Boris Choy, Minh-Ngoc Tran, Junbin Gao

    Commodity demand forecasting is challenging because market demand is shaped by noisy macroeconomic indicators, delayed cross-feature transmissions, regime-sensitive interactions, and economic restrictions such as non-positive direct price elasticity. Existing deep forecasting models provide strong nonlinear capacity, but they often lack mechanisms for controlling signal noise, stabilizing temporal propagation, exposing dynamic market relations, and ruling out economically implausible responses. We propose an economically disciplined operator learning framework for stable commodity demand forecasting. The key idea is to learn the forecasting map as a constrained composition of operators rather than as an unrestricted black-box regression function. The framework uses a fixed Chebyshev spectral projection to retain auditable low-order temporal content, a contraction-oriented state-space-inspired encoder to obtain stable temporal representations, a dynamic sparse cross-feature graph with bounded diffusion to model market transmission, and an elasticity-constrained decoder to enforce or penalize positive price-response violations. We establish theoretical guarantees for projection optimality, local perturbation stability, and non-positive direct price response. Experiments on four commodity demand datasets demonstrate consistent improvements over classical econometric baselines, transparent distributed-lag regressions, recurrent networks, Transformer-based forecasters, and recent long-context architectures. Further ablation and sensitivity analyses show that the performance gains come from the unified design principle of stable, sparse, and economically admissible operator learning.

  • Open Access

    Importance Weighted Variational Inference without the Reparameterization Trick

    by Kamélia Daudel, Minh-Ngoc Tran, Cheng Zhang

    Importance weighted variational inference (VI) approximates densities known up to a normalizing constant by optimizing bounds that tighten with the number of Monte Carlo samples $N$. Standard optimization relies on reparameterized gradient estimators, which are well-studied theoretically yet restrict both the choice of the data-generating process and the variational approximation. While REINFORCE gradient estimators do not suffer from such restrictions, they lack rigorous theoretical justification. In this paper, we provide the first comprehensive analysis of REINFORCE gradient estimators in importance weighted VI, leveraging this theoretical foundation to diagnose and resolve fundamental deficiencies in current state-of-the-art estimators. Specifically, we introduce and examine a generalized family of variational inference for Monte Carlo objectives (VIMCO) gradient estimators. We prove that state-of-the-art VIMCO gradient estimators exhibit a vanishing signal-to-noise ratio (SNR) as $N$ increases, which prevents effective optimization. To overcome this issue, we propose the novel VIMCO-$\star$ gradient estimator and show that it averts the SNR collapse of existing VIMCO gradient estimators by achieving a $\sqrt{N}$ SNR scaling instead. We demonstrate its superior empirical performance compared to current VIMCO implementations in challenging settings where reparameterized gradients are typically unavailable.

  • Loss-Based Bayesian Sequential Prediction of Value-at-Risk with a Long-Memory and Non-Linear Realized Volatility Model

    by Rangika Peiris, Minh-Ngoc Tran, Chao Wang, Richard Gerlach

    A long-memory and non-linear realized volatility model class is proposed for direct Value-at-Risk (VaR) forecasting. This model, referred to as RNN-HAR, extends the heterogeneous autoregressive (HAR) model, a framework known for efficiently capturing long memory in realized measures, by integrating a Recurrent Neural Network (RNN) to handle the non-linear dynamics. Quantile loss-based generalized Bayesian method with Sequential Monte Carlo is employed for model estimation and sequential prediction in RNN-HAR. The empirical analysis is conducted using daily closing prices and realized measures with around 12 years of data till 2022, covering 31 market indices. The proposed model’s one-step-ahead VaR forecasting performance is compared against a basic HAR model and its extensions. The results demonstrate that the proposed RNN-HAR model consistently outperforms all other models considered in the study. The implementation code of the HAR-RNN model is publicly available on GitHub: https://github.com/chaowang-usyd/RNN-HAR.


  • Open Access

    Variational Bayesian inference for models with nuisance parameters and an intractable likelihood

    by Y. H. Paco Tseng, Minh-Ngoc Tran, Robert Kohn

    A primary challenge in Bayesian analysis lies in computing the posterior distribution of model parameters, especially for models with a large number of parameters or intractable likelihoods. Often, the focus is on a subset of parameters, with the remainder regarded as nuisance parameters introduced for computational convenience. This complexity necessitates refined computational methods. Variational Bayesian inference (VB) has emerged as a powerful solution, enhancing computational efficiency by recasting inference as an optimization problem within a family of tractable distributions. However, common VB techniques sometimes fall short, especially for models with nuisance parameters or intractable likelihoods. After identifying characteristics of suboptimal VB methods, we build upon the Hybrid Variational Bayes (HVB) approach introduced by Loaiza-Maya et al. (2022) and develop an extended and unified HVB framework designed to achieve more precise Bayesian inference in such scenarios. Through theoretical exploration and a series of illustrative examples, our approach demonstrates notable improvements over traditional VB methods.


  • Open Access

    A long short-term memory enhanced realized conditional heteroskedasticity model

    by Chen Liu, Chao Wang, Minh-Ngoc Tran, Robert Kohn


  • Open Access

    Deep learning enhanced volatility modeling with covariates

    by Hien Thi Nguyen, Hoang Nguyen, Minh-Ngoc Tran


  • Open Access

    Natural Gradient Variational Bayes without Fisher Matrix Analytic Calculation and Its Inversion

    by A. Godichon-Baggioni, D. Nguyen, M-N Tran

    This paper introduces a method for efficiently approximating the inverse of the Fisher information matrix, a crucial step in achieving effective variational Bayes inference. A notable aspect of our approach is the avoidance of analytically computing the Fisher information matrix and its explicit inversion. Instead, we introduce an iterative procedure for generating a sequence of matrices that converge to the inverse of Fisher information. The natural gradient variational Bayes algorithm without analytic expression of the Fisher matrix and its inversion is provably convergent and achieves a convergence rate of order O(log s/s), with s the number of iterations. We also obtain a central limit theorem for the iterates. Implementation of our method does not require storage of large matrices, and achieves a linear complexity in the number of variational parameters. Our algorithm exhibits versatility, making it applicable across a diverse array of variational Bayes domains, including Gaussian approximation and normalizing flow Variational Bayes. We offer a range of numerical examples to demonstrate the efficiency and reliability of the proposed variational Bayes method.

  • Efficient selection between hierarchical cognitive models: Cross-validation with variational Bayes.

    by Viet Hung Dao, David Gunawan, Minh-Ngoc Tran, Robert Kohn, Guy E. Hawkins, Scott D. Brown

    Model comparison is the cornerstone of theoretical progress in psychological research. Common practice overwhelmingly relies on tools that evaluate competing models by balancing in-sample descriptive adequacy against model flexibility, with modern approaches advocating the use of marginal likelihood for hierarchical cognitive models. Cross-validation is another popular approach but its implementation remains out of reach for cognitive models evaluated in a Bayesian hierarchical framework, with the major hurdle being its prohibitive computational cost. To address this issue, we develop novel algorithms that make variational Bayes (VB) inference for hierarchical models feasible and computationally efficient for complex cognitive models of substantive theoretical interest. It is well known that VB produces good estimates of the first moments of the parameters, which gives good predictive densities estimates. We thus develop a novel VB algorithm with Bayesian prediction as a tool to perform model comparison by cross-validation, which we refer to as CVVB. In particular, CVVB can be used as a model screening device that quickly identifies bad models. We demonstrate the utility of CVVB by revisiting a classic question in decision making research: what latent components of processing drive the ubiquitous speed-accuracy tradeoff? We demonstrate that CVVB strongly agrees with model comparison via marginal likelihood, yet achieves the outcome in much less time. Our approach brings cross-validation within reach of theoretically important psychological models, making it feasible to compare much larger families of hierarchically specified cognitive models than has previously been possible. To enhance the applicability of the algorithm, we provide Matlab code together with a user manual so users can easily implement VB and/or CVVB for the models considered in this article and their variants. (PsycInfo Database Record (c) 2022 APA, all rights reserved).