Zobrazeno 1 - 10
of 262
pro vyhledávání: '"Wang, Ziqiao"'
Current image-text retrieval methods have demonstrated impressive performance in recent years. However, they still face two problems: the inter-modal matching missing problem and the intra-modal semantic loss problem. These problems can significantly
Externí odkaz:
http://arxiv.org/abs/2403.05261
Autor:
Wang, Ziqiao, Mao, Yongyi
Unsupervised domain adaptation (UDA) plays a crucial role in addressing distribution shifts in machine learning. In this work, we improve the theoretical foundations of UDA proposed by Acuna et al. (2021) by refining their f-divergence-based discrepa
Externí odkaz:
http://arxiv.org/abs/2402.01887
Publikováno v:
Rep. Prog. Phys. 87 (2024) 014502
Superconductor-insulator/metal transition (SIT/SMT) as a paradigm of quantum phase transition has been a research highlight over the last three decades. Benefit from recent developments in the fabrication and measurements of 2D superconducting films
Externí odkaz:
http://arxiv.org/abs/2312.07354
Autor:
Qi, Shichao, Liu, Yi, Wang, Ziqiao, Chen, Fucong, Li, Qian, Ji, Haoran, Li, Rao, Li, Yanan, Fang, Jingchao, Liu, Haiwen, Wang, Fa, Jin, Kui, Xie, X. C., Wang, Jian
Disorder is ubiquitous in real materials and can have dramatic effects on quantum phase transitions. Originating from the disorder enhanced quantum fluctuation, quantum Griffiths singularity (QGS) has been revealed as a universal phenomenon in quantu
Externí odkaz:
http://arxiv.org/abs/2311.06710
Autor:
Wang, Ziqiao, Mao, Yongyi
We present new information-theoretic generalization guarantees through the a novel construction of the "neighboring-hypothesis" matrix and a new family of stability notions termed sample-conditioned hypothesis (SCH) stability. Our approach yields sha
Externí odkaz:
http://arxiv.org/abs/2310.20102
Mixup, which creates synthetic training instances by linearly interpolating random sample pairs, is a simple and yet effective regularization technique to boost the performance of deep models trained with SGD. In this work, we report a previously uno
Externí odkaz:
http://arxiv.org/abs/2303.01475
Autor:
Wang, Ziqiao, Mao, Yongyi
In this work, we present a variety of novel information-theoretic generalization bounds for learning algorithms, from the supersample setting of Steinke & Zakynthinou (2020)-the setting of the "conditional mutual information" framework. Our developme
Externí odkaz:
http://arxiv.org/abs/2302.02432
Autor:
Wang, Ziqiao
Forests cover 30% of the Earth’s surface area, but most of the tree canopy reaches up to tens of meters above the ground, making it challenging to explore, and thus our knowledge of it is very limited. To help scientists further grasp the biologica
Externí odkaz:
http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-321088
Autor:
Wang, Ziqiao, Mao, Yongyi
Stochastic differential equations (SDEs) have been shown recently to characterize well the dynamics of training machine learning models with SGD. When the generalization error of the SDE approximation closely aligns with that of SGD in expectation, i
Externí odkaz:
http://arxiv.org/abs/2211.10691
Autor:
Wang, Ziqiao, Mao, Yongyi
This paper uses information-theoretic tools to analyze the generalization error in unsupervised domain adaptation (UDA). We present novel upper bounds for two notions of generalization errors. The first notion measures the gap between the population
Externí odkaz:
http://arxiv.org/abs/2210.00706