Filters








2,219 Hits in 5.2 sec

Federated Learning: Balancing the Thin Line Between Data Intelligence and Privacy [article]

Sherin Mary Mathews, Samuel A. Assefa
2022 arXiv   pre-print
The work also presents an overview of current training challenges for federated learning, focusing on handling non-i.i.d. data, high dimensionality issues, and heterogeneous architecture, and discusses  ...  Federated learning holds great promise in learning from fragmented sensitive data and has revolutionized how machine learning models are trained.  ...  Horizontal federated learning is a federated learning approach in which datasets on the devices share the same attributes in different instances (Yang et al. 2019) .  ... 
arXiv:2204.13697v1 fatcat:rvlsrnk66jblzguy2vnh3thgtu

A Comprehensive Survey on Trustworthy Graph Neural Networks: Privacy, Robustness, Fairness, and Explainability [article]

Enyan Dai, Tianxiang Zhao, Huaisheng Zhu, Junjie Xu, Zhimeng Guo, Hui Liu, Jiliang Tang, Suhang Wang
2022 arXiv   pre-print
Consequently, trustworthy GNNs in various aspects are emerging to prevent the harm from GNN models and increase the users' trust in GNNs.  ...  leak private information, are vulnerable to adversarial attacks, can inherit and magnify societal bias from training data and lack interpretability, which have risk of causing unintentional harm to the users  ...  They view the training on a client as a task in meta-learning and learn a global model to mitigate the issue of non-i.i.d data.  ... 
arXiv:2204.08570v1 fatcat:7c3pkxitrbhgxj6fytn6f3r644

Algorithm Fairness in AI for Medicine and Healthcare [article]

Richard J. Chen, Tiffany Y. Chen, Jana Lipkova, Judy J. Wang, Drew F.K. Williamson, Ming Y. Lu, Sharifa Sahai, Faisal Mahmood
2022 arXiv   pre-print
In this perspective article, we summarize the intersectional field of fairness in machine learning through the context of current issues in healthcare, outline how algorithmic biases (e.g. - image acquisition  ...  Lastly, we also review emerging technology for mitigating bias via federated learning, disentanglement, and model explainability, and their role in AI-SaMD development.  ...  In application to federated learning and medical imaging, frameworks such as FedDis has been demonstrated to isolate sensitive attributes in non-i.i.d.  ... 
arXiv:2110.00603v2 fatcat:pspb6bqqxjh45an5mhqohysswu

FederatedScope-GNN: Towards a Unified, Comprehensive and Efficient Package for Federated Graph Learning [article]

Zhen Wang, Weirui Kuang, Yuexiang Xie, Liuyi Yao, Yaliang Li, Bolin Ding, Jingren Zhou
2022 arXiv   pre-print
However, federated graph learning (FGL), even though graph data are prevalent, has not been well supported due to its unique characteristics and requirements.  ...  The incredible development of federated learning (FL) has benefited various tasks in the domains of computer vision and natural language processing, and the existing frameworks such as TFF and FATE has  ...  non-i.i.d i.i.d. (a) Monitoring the FGL course on i.i.d. and non-i.i.d. datasets constructed by FedcSBM. An example of personalizing GNN: Each client has its dedicated encoder and decoder.  ... 
arXiv:2204.05562v5 fatcat:mlpz7xt5c5ckrojtsc52etcwmm

Incentivising Exploration and Recommendations for Contextual Bandits with Payments [article]

Priyank Agrawal, Theja Tulabandhula
2020 arXiv   pre-print
We propose a contextual bandit based model to capture the learning and social welfare goals of a web platform in the presence of myopic users.  ...  By using payments to incentivize these agents to explore different items/recommendations, we show how the platform can learn the inherent attributes of items and achieve a sublinear regret while maximizing  ...  Essentially we identify a way to adapt and extend the non principalagent setting of [11] to our platform-user interaction model.  ... 
arXiv:2001.07853v1 fatcat:7ed2qdnxrze2bmna7ezzijo25e

Transmit Antenna Selection for Massive MIMO-GSM with Machine Learning [article]

Selen Gecgel, Caner Goztepe, Gunes Karabulut Kurt
2019 arXiv   pre-print
Both decision tree and multi-layer perceptrons approaches are adopted for the GSM transmitter.  ...  The observations are validated through measurement results over the designed 16× 4 MIMO test-bed using software defined radio nodes.  ...  Perceptrons: MLP is an artificial neural network structure and has the ability to learn non-linear functions via a given set of features.  ... 
arXiv:1903.04460v1 fatcat:zbdlav3d6bdtzbeqxx2yv6lhti

Federated and Transfer Learning: A Survey on Adversaries and Defense Mechanisms [article]

Ehsan Hallaji, Roozbeh Razavi-Far, Mehrdad Saif
2022 arXiv   pre-print
The advent of federated learning has facilitated large-scale data exchange amongst machine learning models while maintaining privacy.  ...  One of the most significant advancements in this domain is the incorporation of transfer learning into federated learning, which overcomes fundamental constraints of primary federated learning, particularly  ...  . data and its effectiveness in non-i.i.d. scenarios is unknown.  ... 
arXiv:2207.02337v1 fatcat:rf4fdiunnnehjpvjhbmncrt3ka

Multi-fairness under class-imbalance [article]

Arjun Roy, Vasileios Iosifidis, Eirini Ntoutsi
2022 arXiv   pre-print
To this end, we introduce a new fairness measure, Multi-Max Mistreatment (MMM), which considers both (multi-attribute) protected group and class membership of instances to measure discrimination.  ...  Recent studies showed that datasets used in fairness-aware machine learning for multiple protected attributes (referred to as multi-discrimination hereafter) are often imbalanced.  ...  Basics and Multi-Max Mistreatment(MMM ) fairness We assume a dataset D = (u (i) , s (i) , y (i) ) ∼ P of n instances drawn from the i.i.d distribution P over the domain U × S × Y , where U is the subspace  ... 
arXiv:2104.13312v3 fatcat:gu6jq2g7avbutlpns75qdijfia

On the relation between multi-instance learning and semi-supervised learning

Zhi-Hua Zhou, Jun-Ming Xu
2007 Proceedings of the 24th international conference on Machine learning - ICML '07  
Multi-instance learning and semi-supervised learning are different branches of machine learning.  ...  In this paper, we establish a bridge between these two branches by showing that multi-instance learning can be viewed as a special case of semi-supervised learning.  ...  ., 1997) , we think multi-instance learning relaxes the i.i.d. assumption made by traditional supervised learning.  ... 
doi:10.1145/1273496.1273643 dblp:conf/icml/ZhouX07 fatcat:7d5n5b7bijh4vdo4ghdj3gci3u

CAPE: Context-Aware Private Embeddings for Private Language Learning [article]

Richard Plant, Dimitra Gkatzia, Valerio Giuffrida
2021 arXiv   pre-print
Obtaining text representations or embeddings using these models presents the possibility of encoding personally identifiable information learned from language and context cues that may present a risk to  ...  In addition, CAPE employs an adversarial training regime that obscures identified private variables.  ...  Even seemingly innocuous collections of metadata (Xu et al., 2008) such as data provided by the users (e.g. at registration time on social media) or data which has been cleansed of identifying attributes  ... 
arXiv:2108.12318v1 fatcat:qs254y4bdvhejhpk4uuaz33uva

Guest Editorial Special Issue on Emerging Computational Intelligence Techniques for Decision Making With Big Data in Uncertain Environments

Weiping Dingr, Nikhil R. Pal, Chin-Teng Lin, Yiu-ming Cheung, Zehong Cao, Wenjian Luo
2021 IEEE Transactions on Emerging Topics in Computational Intelligence  
The utility is computed by allowing users to weigh the course related attributes according to their preferences.  ...  The proposed online learning algorithms could be well applied to multi-agent decision making based on big data in an unknown environment.  ... 
doi:10.1109/tetci.2021.3049701 fatcat:fwz2kgi3nnbgvlbednhwelq23i

Voting-based Approaches For Differentially Private Federated Learning [article]

Yuqing Zhu, Xiang Yu, Yi-Hsuan Tsai, Francesco Pittaluga, Masoud Faraki, Manmohan chandraker, Yu-Xiang Wang
2021 arXiv   pre-print
In this work, inspired by knowledge transfer non-federated privacy learning from Papernot et al.(2017; 2018), we design two new DPFL schemes, by voting among the data labels returned from each local model  ...  Differentially Private Federated Learning (DPFL) is an emerging field with many applications.  ...  MNIST Dataset with Non-I.I.D Partition: In both CelebA and Digit experiments, we I.I.D partition each dataset into different agents.  ... 
arXiv:2010.04851v2 fatcat:ff5qqlgdonhefexhjoc5fep32q

Extracting Social Dimensions Using Fiedler Embedding

Xi Wang, Gita Sukthankar
2011 2011 IEEE Third Int'l Conference on Privacy, Security, Risk and Trust and 2011 IEEE Third Int'l Conference on Social Computing  
The network structure causes these data instances to no longer remain independently identically distributed (i.i.d.).  ...  Networked data, such as data from social media, contains instances of multiple types that are related through different types of links.  ...  The network structure makes the data instances no longer independently identically distributed (i.i.d.).  ... 
doi:10.1109/passat/socialcom.2011.144 dblp:conf/socialcom/WangS11 fatcat:xfhf2alj5bgjvntwc6pg2rx6ge

Collective Semi-Supervised Learning for User Profiling in Social Media [article]

Richard J. Oentaryo, Ee-Peng Lim, Freddy Chong Tat Chua, Jia-Wei Low, David Lo
2016 arXiv   pre-print
The joint learning from multiple relationships and unlabeled data yields a computationally sound and accurate approach to model user attributes in social media.  ...  The abundance of user-generated data in social media has incentivized the development of methods to infer the latent attributes of users, which are crucially useful for personalization, advertising and  ...  RELATED WORK We first survey related works on user attribute profiling, semi-supervised learning, and multi-relational learning. We then discuss how our approach differs from these works.  ... 
arXiv:1606.07707v1 fatcat:2dpz6hiruzh5logx3oh3qpvt7q

SpecAE: Spectral AutoEncoder for Anomaly Detection in Attributed Networks [article]

Yuening Li, Xiao Huang, Jundong Li, Mengnan Du, Na Zou
2019 arXiv   pre-print
While most existing algorithms assume that instances are i.i.d., in many practical scenarios, links describing instance-to-instance dependencies and interactions are available.  ...  Such systems are called attributed networks.  ...  Introduction Anomaly detection targets at identifying the rare instances that behave significantly different from the majority instances.  ... 
arXiv:1908.03849v3 fatcat:qdhuhlfyffdnjiaxlrx2izpocy
« Previous Showing results 1 — 15 out of 2,219 results