Výsledky vyhledávání - "Computing methodologies Machine learning"

Muffin: A Framework Toward Multi-Dimension AI Fairness by Uniting Off-the-Shelf Models Autor Sheng, Yi, Yang, Junhuan, Yang, Lei, Shi, Yiyu, Hu, Jingtong, Jiang, Weiwen

“…Model fairness (a.k.a., bias) has become one of the most critical problems in a wide range of AI applications. An unfair model in autonomous driving may cause…”
Získat plný text

PrefixRL: Optimization of Parallel Prefix Circuits using Deep Reinforcement Learning Autor Roy, Rajarshi, Raiman, Jonathan, Kant, Neel, Elkin, Ilyas, Kirby, Robert, Siu, Michael, Oberman, Stuart, Godil, Saad, Catanzaro, Bryan

“…In this work, we present a reinforcement learning (RL) based approach to designing parallel prefix circuits such as adders or priority encoders that are…”
Získat plný text

Code Difference Guided Adversarial Example Generation for Deep Code Models Autor Tian, Zhao, Chen, Junjie, Jin, Zhi

“…Adversarial examples are important to test and enhance the robustness of deep code models. As source code is discrete and has to strictly stick to complex…”
Získat plný text

Toward Individual Fairness Testing with Data Validity Autor Kitamura, Takashi, Amasaki, Sousuke, Inoue, Jun, Isobe, Yoshinao, Toda, Takahisa

“…Individual fairness testing (Ift) is a framework to find discriminatory instances within a given classifier. In this paper, we show our idea of a Ift…”
Získat plný text

Softermax: Hardware/Software Co-Design of an Efficient Softmax for Transformers Autor Stevens, Jacob R., Venkatesan, Rangharajan, Dai, Steve, Khailany, Brucek, Raghunathan, Anand

“…Transformers have transformed the field of natural language processing. Their superior performance is largely attributed to the use of stacked "self-attention"…”
Získat plný text

Enabling On-Device Self-Supervised Contrastive Learning with Selective Data Contrast Autor Wu, Yawen, Wang, Zhepeng, Zeng, Dewen, Shi, Yiyu, Hu, Jingtong

“…After a model is deployed on edge devices, it is desirable for these devices to learn from unlabeled data to continuously improve accuracy. Contrastive…”
Získat plný text

PreSto: An In-Storage Data Preprocessing System for Training Recommendation Models Autor Lee, Yunjae, Kim, Hyeseong, Rhu, Minsoo

“…Training recommendation systems (RecSys) faces several challenges as it requires the "data preprocessing" stage to preprocess an ample amount of raw data and…”
Získat plný text

DistHD: A Learner-Aware Dynamic Encoding Method for Hyperdimensional Classification Autor Wang, Junyao, Huang, Sitao, Imani, Mohsen

“…The Internet of Things (IoT) has become an emerging trend that connects heterogeneous devices and enables them with new capabilities. Many applications exploit…”
Získat plný text

UniGenCoder: Merging SEQ2SEQ and SEQ2TREE Paradigms for Unified Code Generation Autor Shao, Liangying, Yan, Yanfu, Poshyvanyk, Denys, Su, Jinsong

“…Deep learning-based code generation has completely transformed the way developers write programs today. Existing approaches to code generation have focused…”
Získat plný text

Fast Adversarial Training with Dynamic Batch-level Attack Control Autor Jung, Jaewon, Song, Jaeyong, Jang, Hongsun, Lee, Hyeyoon, Choi, Kanghyun, Park, Noseong, Lee, Jinho

“…Despite the fact that adversarial training provides an effective protection against adversarial attacks, it suffers from a huge computational overhead. To…”
Získat plný text

DARL: Distributed Reconfigurable Accelerator for Hyperdimensional Reinforcement Learning Autor Chen, Hanning, Issa, Mariam, Ni, Yang, Imani, Mohsen

“…Reinforcement Learning (RL) is a powerful technology to solve decision-making problems such as robotics control. Modern RL algorithms, i.e., Deep Q-Learning,…”
Získat plný text

Robust Tickets Can Transfer Better: Drawing More Transferable Subnetworks in Transfer Learning Autor Fu, Yonggan, Yuan, Ye, Wu, Shang, Yuan, Jiayi, Lin, Yingyan Celine

“…Transfer learning leverages feature representations of deep neural networks (DNNs) pretrained on source tasks with rich data to empower effective finetuning on…”
Získat plný text

CAE-DFKD: Bridging the Transferability Gap in Data-Free Knowledge Distillation Autor Zhang, Zherui, Wang, Changwei, Xu, Rongtao, Xu, Wenhao, Xu, Shibiao, Zhang, Yu, Zhou, Jie, Guo, Li

“…Data-Free Knowledge Distillation (DFKD) enables the knowledge transfer from the given pre-trained teacher network to the target student model without access to…”
Získat plný text

Making Fair ML Software using Trustworthy Explanation Autor Chakraborty, Joymallya, Peng, Kewen, Menzies, Tim

“…Machine learning software is being used in many applications (finance, hiring, admissions, criminal justice) having huge social impact. But sometimes the…”
Získat plný text

Dissecting Global Search: A Simple Yet Effective Method to Boost Individual Discrimination Testing and Repair Autor Quan, Lili, Li, Tianlin, Xie, Xiaofei, Chen, Zhenpeng, Chen, Sen, Jiang, Lingxiao, Li, Xiaohong

“…Deep Learning (DL) has achieved significant success in socially critical decision-making applications but often exhibits unfair behaviors, raising social…”
Získat plný text

Iterative Generation of Adversarial Example for Deep Code Models Autor Huang, Li, Sun, Weifeng, Yan, Meng

“…Deep code models are vulnerable to adversarial attacks, making it possible for semantically identical inputs to trigger different responses. Current black-box…”
Získat plný text

Fairquant: Certifying and Quantifying Fairness of Deep Neural Networks Autor Kim, Brian Hyeongseok, Wang, Jingbo, Wang, Chao

“…We propose a method for formally certifying and quantifying individual fairness of deep neural networks (DNN). Individual fairness guarantees that any two…”
Získat plný text

On-the-fly Improving Performance of Deep Code Models via Input Denoising Autor Tian, Zhao, Chen, Junjie, Zhang, Xiangyu

“…Deep learning has been widely adopted to tackle various code-based tasks by building deep code models based on a large amount of code snippets. While these…”
Získat plný text

Code Prediction by Feeding Trees to Transformers Autor Kim, Seohyun, Zhao, Jinman, Tian, Yuchi, Chandra, Satish

“…Code prediction, more specifically autocomplete, has become an essential feature in modern IDEs. Autocomplete is more effective when the desired next token is…”
Získat plný text

Diversity Drives Fairness: Ensemble of Higher Order Mutants for Intersectional Fairness of Machine Learning Software Autor Chen, Zhenpeng, Li, Xinyue, Zhang, Jie M., Sarro, Federica, Liu, Yang

“…Intersectional fairness is a critical requirement for Machine Learning (ML) software, demanding fairness across subgroups defined by multiple protected…”
Získat plný text