no code implementations • 13 Feb 2024 • Shaeke Salman, Md Montasir Bin Shams, Xiuwen Liu, Lingjiong Zhu
Transformer-based models have dominated natural language processing and other areas in the last few years due to their superior (zero-shot) performance on benchmark datasets.
no code implementations • 28 Jan 2024 • Shaeke Salman, Md Montasir Bin Shams, Xiuwen Liu
Pre-trained large foundation models play a central role in the recent surge of artificial intelligence, resulting in fine-tuned models with remarkable abilities when measured on benchmark datasets, standard exams, and applications.
no code implementations • 16 Dec 2023 • Canlin Zhang, Xiuwen Liu
Embedding-based models usually need fine-tuning on new entity embeddings, and hence are difficult to be directly applied to inductive link prediction tasks.
no code implementations • 9 Oct 2023 • Yuanhang Shao, Xiuwen Liu
However, utilizing labels and features jointly in higher-order graphs has not been explored.
no code implementations • 25 Jan 2023 • Luis Villamil, Ryan Bausback, Shaeke Salman, Ting L. Liu, Conrad Horn, Xiuwen Liu
We further incorporate weight decay, batch normalization, dropout, and label smoothing to improve the generalization of the trained models.
no code implementations • 14 Jan 2023 • John Lazzari, Xiuwen Liu
However, in low dimensional settings, a severe spectral bias occurs that obstructs convergence to high frequency components entirely.
1 code implementation • NAACL 2021 • Daniel Bi{\'s}, Maksim Podkorytov, Xiuwen Liu
The success of language models based on the Transformer architecture appears to be inconsistent with observed anisotropic properties of representations learned by such models.
1 code implementation • bioRxiv 2021 • Yuan Zhang, Arunima Mandal, Kevin Cui, Xiuwen Liu, Jinfeng Zhang
The prediction is very fast compared with other protein sequence prediction servers - it takes only a few minutes for a query protein on average.
1 code implementation • 22 Apr 2020 • Canlin Zhang, Xiuwen Liu
In order to create suitable labels for the training of sense spectra, we designed a new similarity measurement for noun and verb synsets in WordNet.
1 code implementation • 18 Mar 2020 • Canlin Zhang, Xiuwen Liu, Daniel Bis
To improve the generalization of the representations for natural language processing tasks, words are commonly represented using vectors, where distances among the vectors are related to the similarity of the words.
no code implementations • 18 Oct 2019 • Shaeke Salman, Canlin Zhang, Xiuwen Liu, Washington Mio
We show that the generalization intervals of a ReLU network behave similarly along pairwise directions between samples of the same label in both real and random cases on the MNIST and CIFAR-10 datasets.
no code implementations • 14 May 2019 • Shaeke Salman, Seyedeh Neelufar Payrovnaziri, Xiuwen Liu, Pablo Rengifo-Moreno, Zhe He
In particular, while the proposed method maintains similar interpretability as conventional shallow models such as logistic regression, it improves the prediction accuracy significantly.
no code implementations • 3 Feb 2019 • Meysam Ghaffari, Ashok Srinivasan, Xiuwen Liu
We obtained over 90% accuracy for large subsets on a commonly used dataset.
no code implementations • 19 Jan 2019 • Shaeke Salman, Xiuwen Liu
Assisted by the availability of data and high performance computing, deep learning techniques have achieved breakthroughs and surpassed human performance empirically in difficult tasks, including object recognition, speech recognition, and natural language processing.
no code implementations • 2 Aug 2017 • Atharva Sharma, Xiuwen Liu, Xiaojun Yang
Even with the increased number of satellite systems and sensors acquiring data with improved spectral, spatial, radiometric and temporal characteristics and the new data distribution policy, most existing land cover datasets were derived from a pixel-based single-date multi-spectral remotely sensed image with low accuracy.
no code implementations • 22 Dec 2015 • Jiangbo Yuan, Xiuwen Liu
The success of product quantization (PQ) for fast nearest neighbor search depends on the exponentially reduced complexities of both storage and computation with respect to the codebook size.