Enhancing Federated Learning Evaluation: Exploring Instance-Level Insights with SQUARES in Image Classification Models

Deleted Journal, Journal Year: 2024, Volume and Issue: 20(7s), P. 2516 - 2523

Published: May 29, 2024

Federated Learning (FL) presents a novel approach within the domain of Machine (ML)—enabling training ML models in distributed manner. This paradigmatic shift ensures preservation user privacy by conducting process at edge, where data remains localized on individual devices. stands contrast to conventional centralized paradigm—where users transmit central server for processing. The FL framework operates more heterogeneous environment characterized diverse distributions across clients, departure from engineer-analyzed datasets typical paradigms. Whereas, evaluation typically relies statistical techniques such as accuracy, recall, precision, log loss, and confusion matrix, alongside visualization methods. While these provide an overview model's performance utilization—they may lack granularity when comparing with disparate characteristics. However, SQUARES technique, offers nuanced model instance level. facilitates examination biases, outlier detection, behavior during samples. Therefore, this study development image classification various scenarios—utilizing prototype. In addressing intricate scenarios, we aim augment traditional visualizations metrics, thereby uncovering insights nuances that elude standard methods prevalent benchmarks.

Language: Английский

---

Optimizing Photometric Light Curve Analysis: Evaluating Scipy's Minimize Function for Eclipse Mapping of Cataclysmic Variables

Deleted Journal, Journal Year: 2024, Volume and Issue: 20(7s), P. 2557 - 2566

Published: May 29, 2024

With a particular focus on Scipy's minimize function the eclipse mapping method is thoroughly researched and implemented utilizing Python essential libraries. Many optimization techniques are used, including Sequential Least Squares Programming (SLSQP), Nelder-Mead, Conjugate Gradient (CG). However, for purpose of examining photometric light curves these methods seek to solve maximum entropy equation under chi-squared constraint. Therefore, first evaluated two-dimensional Gaussian data without restriction, then they used map accretion disc uncover structure Cataclysmic Variable KIC 201325107. Critical analysis performed code find possible faults design problems. Additionally, shows how several factors impacting computing time image quality included variance in weighting, resolution, number points curve, degree

Language: Английский

---

SPACE TECHNOLOGIES AND DEVOPS: INTEGRATING POSITIONING AND TIMING SYSTEMS FOR RELIABLE SOFTWARE DEPLOYMENT IN AEROSPACE APPLICATIONS

International Journal for Research Publication and Seminars, Journal Year: 2024, Volume and Issue: 15(3), P. 531 - 541

Published: Sept. 28, 2024

The specific topic of investigation in this paper is the integration PNT technologies into DevOps paradigm for aerospace domain with an emphasis on improved robustness and accuracy software releases. paper's goal to show how systems, especially GNSS machine learning applied time-series data, can solve several critical issues faced deployment aerospace, whether using simulation reports real-world case study examples, areas ways which PNT-enabled pipeline may be employed highlighted. fact that timing positioning investigations improve also reduces synchronization errors, as evidenced by research. Through preceding technologies, organizations will able elevate reliability installations, hence positively contributing operational safety aspects space missions. This indicates a significant advancement what must done handle processes requirements applications.

Language: Английский

---

SPACE TECHNOLOGIES AND DEVOPS: INTEGRATING POSITIONING AND TIMING SYSTEMS FOR RELIABLE SOFTWARE DEPLOYMENT IN AEROSPACE APPLICATIONS

International Journal for Research Publication and Seminars, Journal Year: 2024, Volume and Issue: 15(2), P. 286 - 294

Published: June 29, 2024

The specific topic of investigation in this paper is the integration PNT technologies into DevOps paradigm for aerospace domain with an emphasis on improved robustness and accuracy software releases. paper's goal to show how systems, especially GNSS machine learning applied time-series data, can solve several critical issues faced deployment aerospace, whether using simulation reports real-world case study examples, areas ways which PNT-enabled pipeline may be employed highlighted. fact that timing positioning investigations improve also reduces synchronization errors, as evidenced by research. Through preceding technologies, organizations will able elevate reliability installations, hence positively contributing operational safety aspects space missions. This indicates a significant advancement what must done handle processes requirements applications.

Language: Английский

---

SPACE TECHNOLOGIES AND DEVOPS: INTEGRATING POSITIONING AND TIMING SYSTEMS FOR RELIABLE SOFTWARE DEPLOYMENT IN AEROSPACE APPLICATIONS

International Journal for Research Publication and Seminars, Journal Year: 2023, Volume and Issue: 14(4), P. 531 - 541

Published: July 30, 2023

The specific topic of investigation in this paper is the integration PNT technologies into DevOps paradigm for aerospace domain with an emphasis on improved robustness and accuracy software releases. paper's goal to show how systems, especially GNSS machine learning applied time-series data, can solve several critical issues faced deployment aerospace, whether using simulation reports real-world case study examples, areas ways which PNT-enabled pipeline may be employed highlighted. fact that timing positioning investigations improve also reduces synchronization errors, as evidenced by research. Through preceding technologies, organizations will able elevate reliability installations, hence positively contributing operational safety aspects space missions. This indicates a significant advancement what must done handle processes requirements applications.

Language: Английский

---