arXiv subjects

Bao Nguyen

Publications and source records attributed to Bao Nguyen.

26 records · Page 2Linked to original sources

arXiv2024

Cost-Adaptive Recourse Recommendation by Adaptive Preference Elicitation

Algorithmic recourse recommends a cost-efficient action to a subject to reverse an unfavorable machine learning classification decision. Most existing methods in the literature generate recourse under the assumption of complete knowledge about the cost function. In real-world practice, subjects could have distinct preferences, leading to incomplete information about the underlying cost function of the subject. This paper proposes a two-step approach integrating preference learning into the recourse generation problem. In the first step, we design a question-answering framework to refine the confidence set of the Mahalanobis matrix cost of the subject sequentially. Then, we generate recourse by utilizing two methods: gradient-based and graph-based cost-adaptive recourse that ensures validity while considering the whole confidence set of the cost matrix. The numerical evaluation demonstrates the benefits of our approach over state-of-the-art baselines in delivering cost-efficient recourse recommendations.

cs.LG↗

arXiv2024

Bellman Optimal Stepsize Straightening of Flow-Matching Models

Flow matching is a powerful framework for generating high-quality samples in various applications, especially image synthesis. However, the intensive computational demands of these models, especially during the finetuning process and sampling processes, pose significant challenges for low-resource scenarios. This paper introduces Bellman Optimal Stepsize Straightening (BOSS) technique for distilling flow-matching generative models: it aims specifically for a few-step efficient image sampling while adhering to a computational budget constraint. First, this technique involves a dynamic programming algorithm that optimizes the stepsizes of the pretrained network. Then, it refines the velocity network to match the optimal step sizes, aiming to straighten the generation paths. Extensive experimental evaluations across image generation tasks demonstrate the efficacy of BOSS in terms of both resource utilization and image quality. Our results reveal that BOSS achieves substantial gains in efficiency while maintaining competitive sample quality, effectively bridging the gap between low-resource constraints and the demanding requirements of flow-matching generative models. Our paper also fortifies the responsible development of artificial intelligence, offering a more sustainable generative model that reduces computational costs and environmental footprints. Our code can be found at https://github.com/nguyenngocbaocmt02/BOSS.

cs.CV↗

arXiv2023

AI-assisted Learning for Electronic Engineering Courses in High Education

This study evaluates the efficacy of ChatGPT as an AI teaching and learning support tool in an integrated circuit systems course at a higher education institution in an Asian country. Various question types were completed, and ChatGPT responses were assessed to gain valuable insights for further investigation. The objective is to assess ChatGPT's ability to provide insights, personalized support, and interactive learning experiences in engineering education. The study includes the evaluation and reflection of different stakeholders: students, lecturers, and engineers. The findings of this study shed light on the benefits and limitations of ChatGPT as an AI tool, paving the way for innovative learning approaches in technical disciplines. Furthermore, the study contributes to our understanding of how digital transformation is likely to unfold in the education sector.

cs.CY↗

arXiv2023

Multi-dimensional data refining strategy for effective fine-tuning LLMs

Data is a cornerstone for fine-tuning large language models, yet acquiring suitable data remains challenging. Challenges encompassed data scarcity, linguistic diversity, and domain-specific content. This paper presents lessons learned while crawling and refining data tailored for fine-tuning Vietnamese language models. Crafting such a dataset, while accounting for linguistic intricacies and striking a balance between inclusivity and accuracy, demands meticulous planning. Our paper presents a multidimensional strategy including leveraging existing datasets in the English language and developing customized data-crawling scripts with the assistance of generative AI tools. A fine-tuned LLM model for the Vietnamese language, which was produced using resultant datasets, demonstrated good performance while generating Vietnamese news articles from prompts. The study offers practical solutions and guidance for future fine-tuning models in languages like Vietnamese.

cs.CL↗

arXiv2023

JWST CEERS & JADES Active Galaxies at z = 4-7 Violate the Local $M_\bullet-M_\star$ Relation at $>3σ$: Implications for Low-Mass Black Holes and Seeding Models

JWST is revolutionizing our understanding of the high-$z$ Universe by expanding the black hole horizon, looking farther and to smaller masses, and revealing the stellar light of their hosts. By examining JWST galaxies at $z=4-7$ that host H$α$-detected black holes, we investigate (i) the high-$z$ $M_\bullet-M_\star$ relation and (ii) the black hole mass distribution, especially in its low-mass range ($M_\bullet \lesssim 10^{6.5} M_\odot$). With a detailed statistical analysis, our findings conclusively reveal a high-$z$ $M_\bullet-M_\star$ relation that deviates at $>3σ$ confidence level from the local relation. The high-$z$ relation is: $\log(M_\bullet/M_\odot) = -2.43^{+0.83}_{-0.83} + 1.06^{+0.09}_{-0.09} \log(M_\star/M_\odot)$. Black holes are overmassive by $\sim 10-100\times$ compared to their low-$z$ counterparts in galactic hosts of the same stellar mass. This fact is not due to a selection effect in surveys. Moreover, our analysis predicts the possibility of detecting in high-$z$ JWST surveys $5-15\times$ more black holes with $M_\bullet \lesssim 10^{6.5} M_\odot$, and $10-30\times$ more with $M_\bullet \lesssim 10^{8.5} M_\odot$, compared to local relation's predictions. The lighter black holes preferentially occupy galaxies with a stellar mass of $\sim 10^{7.5}-10^8 M_\odot$. We have yet to detect these sources because (i) they may be inactive (duty cycles $1\%-10\%$), (ii) the host overshines the AGN, or (iii) the AGN is obscured and not immediately recognizable by line diagnostics. A search of low-mass black holes in existing JWST surveys will further test the $M_\bullet-M_\star$ relation. Current JWST fields represent a treasure trove of black hole systems at $z = 4-7$; their detection will provide crucial insights into their early evolution and co-evolution with their galactic hosts.

astro-ph.GA↗

arXiv2023

Shadow Geometry of Kerr Naked Singularities

Direct imaging of supermassive black holes (SMBHs) at event horizon scale resolutions, as recently done by the Event Horizon Telescope, allows for testing alternative models to SMBHs such as Kerr naked singularities (KNSs). We demonstrate that the KNS shadow can be closed, open, or vanishing, depending on the spins and observational inclination angles. We study the critical parameters where the KNS shadow opens a gap, a distinctive phenomenon that does not happen with the black hole shadow. We show that the KNS shadow can only be closed for dimensionless spin $a \lesssim 1.18$ and vanishing for $a \gtrsim 1.18$ for certain ranges of inclination angles. We further analyze the effective angular momentum of photon orbits to demonstrate the fundamental connections between light geodesics and the KNS shadow geometry. We also perform numerical general relativistic ray tracing calculations, which reproduce the analytical topological change in the KNS shadow and illustrate other observational features within the shadow due to the lack of an event horizon. By comparing with black hole shadow observations, the topological change in the shadow of KNSs can be used to test the cosmic censorship hypothesis and KNSs as alternative models to SMBHs.

astro-ph.HE↗

arXiv2023

Efficient Failure Pattern Identification of Predictive Algorithms

Given a (machine learning) classifier and a collection of unlabeled data, how can we efficiently identify misclassification patterns presented in this dataset? To address this problem, we propose a human-machine collaborative framework that consists of a team of human annotators and a sequential recommendation algorithm. The recommendation algorithm is conceptualized as a stochastic sampler that, in each round, queries the annotators a subset of samples for their true labels and obtains the feedback information on whether the samples are misclassified. The sampling mechanism needs to balance between discovering new patterns of misclassification (exploration) and confirming the potential patterns of classification (exploitation). We construct a determinantal point process, whose intensity balances the exploration-exploitation trade-off through the weighted update of the posterior at each round to form the generator of the stochastic sampler. The numerical results empirically demonstrate the competitive performance of our framework on multiple datasets at various signal-to-noise ratios.

cs.LG↗

arXiv2020

Unsupervised Region-based Anomaly Detection in Brain MRI with Adversarial Image Inpainting

Medical segmentation is performed to determine the bounds of regions of interest (ROI) prior to surgery. By allowing the study of growth, structure, and behaviour of the ROI in the planning phase, critical information can be obtained, increasing the likelihood of a successful operation. Usually, segmentations are performed manually or via machine learning methods trained on manual annotations. In contrast, this paper proposes a fully automatic, unsupervised inpainting-based brain tumour segmentation system for T1-weighted MRI. First, a deep convolutional neural network (DCNN) is trained to reconstruct missing healthy brain regions. Then, upon application, anomalous regions are determined by identifying areas of highest reconstruction loss. Finally, superpixel segmentation is performed to segment those regions. We show the proposed system is able to segment various sized and abstract tumours and achieves a mean and standard deviation Dice score of 0.771 and 0.176, respectively.

eess.IV↗