Smart devices with built-in sensors, computational capabilities, and network connectivity have become increasingly pervasive. Crowds of smart devices offer opportunities to collectively sense and perform computing tasks at an unprecedented scale. This paper presents Crowd-ML, a privacy-preserving machine learning framework for a crowd of smart devices… (More)
Fig. 4: Comparison of test error for centralized, crowd, and decentralized learning approaches, without delay or privacy consideration. The curves show how error decreases as the number of iterations (number of samples used) increases over time. The batch algorithm is not incremental and therefore is a constant.