https://cis.temple.edu/~jiewu/research/publications/Publication_files/FedCPD.pdf
Prototype Alignment and Contrastive Learning for Im-proved Generalization. In the context of federated learn-ing, the data between clients is typically non-IID.
https://www.fox.temple.edu/directory/detmar-william-straub-tug28766
Information and Management, 57 (7). doi: 10.1016/j.im.2020.103365. Wei, Y., McIntyre, F., & Straub, D. (2020). Does Micro-Blogging Lead to a More Positive Attitude Toward a Brand?—A Perspective of Cultivation Theory. Journal of Promotion Management, 26 (4), 504-523. doi: 10.1080/10496491.2020.1719957. Vance, A., Siponen, M., & Straub, D. (2020).
https://cis.temple.edu/~jiewu/research/publications/Publication_files/FedHAN.pdf
Even upon identi-fying malicious clients, the global model may al-ready be significantly damaged, requiring effec-tive recovery strategies to reduce the attacker’s im-pact. Current recovery methods, which are based on historical update records, are limited in en-vironments with device heterogeneity and asyn-chronous communication.
https://education.temple.edu/help/microsoft
Microsoft Office 365 is available for current Temple students, faculty and staff for free. The software is available for download on up to five PCs or Macs via the Microsoft Download link under TUapplications on TUportal. Along with all the familiar tools, like Word, Excel, and PowerPoint, your Temple license includes access to Microsoft's powerful sharing and collaboration tools: Outlook ...
https://cis.temple.edu/~yu/research/HearBP-info24.pdf
Motivated by the above limitations, we design and im-plement a convenient, comfortable, and continuous in-ear BP monitoring system, HearBP, through an acoustic sensing scheme based on heart sounds. The key inspiration for the design comes from the following aspects.
https://cis.temple.edu/~yu/research/PPGSpotter-info24.pdf
Day Fig. 15. Impact of training data size. tions of the right volar wrist (S1: radial zone, S2: center zone, S3: ulnar zone). As shown in Fig 14, our proposed im-age calibration algorithm emonstrates consistent performance across different sensor positions on the wrist. Specifically, the workload estimation REMS remains lower than 0.65kg regard-l
https://cis.temple.edu/~jiewu/research/publications/Publication_files/INFOCOM2024_PSFL%20Parallel-Sequential%20Federated%20Learning%20with%20Convergence%20Guarantees.pdf
[7] E. Rizk, S. Vlaski, and A. H. Sayed, “Federated learning under im-portance sampling,” IEEE Transactions on Signal Processing, vol. 70, pp. 5381–5396, 2022.