Federated Learning Techniques for Privacy Enhanced Data Mining
Abstract
As data from various sources accumulates, there is an increasing need to protect sensitive information, especially with advancements in technology and the rapid emergence of Artificial Intelligence (AI). Federated learning (FL) has been widely applied in Machine Learning (ML) and Deep Learning (DL), but its role in enhancing privacy in Data Mining (DM) tasks remains significant. This chapter explores how privacy is incorporated into DM using FL techniques. The focus is on privacy methods that have evolved from anonymization, randomization, and differential privacy to more recent FL approaches. Additionally, this review discusses different types of FL, including the Centralized Federated model, which aggregates model details but faces challenges like communication bottlenecks and server limitations. These challenges led to the Decentralized Federated Learning (DFL) approach, which effectively addresses these issues. We also review the metrics used in FL and examine applications where FL can enhance DM processes. By integrating FL into DM, this chapter demonstrates how privacy and efficiency can coexist, making it a promising area for future research.
Read the paper
Related studies
- Active Learning during Federated Learning for Object Detection
- Fairness-aware federated learning
- Decentralized Federated Learning
- Building a Personally Identifiable Information Recognizer in a Privacy Preserved Manner Using Automated Annotation and Federated Learning
- Preservation of Patient Level Privacy: Federated Classification and Calibration Models
- Efficient explainable federated learning
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.