Mobile sensing appears as a promising solution for health inference problem
(e.g., influenza-like symptom recognition) by leveraging diverse smart sensors
to capture fine-grained information about human behaviors and ambient contexts.
Centralized training of machine learning models can place mobile users'
sensitive information under privacy risks due to data breach and
misexploitation. Federated Learning (FL) enables mobile devices to
collaboratively learn global models without the exposure of local private data.
However, there are challenges of on-device FL deployment using mobile sensing:
1) long-term and continuously collected mobile sensing data may exhibit domain
shifts as sensing objects (e.g. humans) have varying behaviors as a result of
internal and/or external stimulus; 2) model retraining using all available data
may increase computation and memory burden; and 3) the sparsity of annotated
crowd-sourced data causes supervised FL to lack robustness. In this work, we
propose FedMobile, an incremental semi-supervised federated learning algorithm,
to train models semi-supervisedly and incrementally in a decentralized online
fashion. We evaluate FedMobile using a real-world mobile sensing dataset for
influenza-like symptom recognition. Our empirical results show that
FedMobile-trained models achieve the best results in comparison to the selected
baseline methods.

移动感知技术结合联邦学习能在保护用户隐私的同时，有效识别类似流感症状，并通过增量半监督联邦学习算法在分散的在线环境中训练模型，取得最佳结果。

基于移动感知的增量半监督联邦学习进行健康推测

Incremental Semi-supervised Federated Learning for Health Inference via  Mobile Sensing

This work considers the problem of decentralized online learning, where the
goal is to track the optimum of the sum of time-varying functions, distributed
across several nodes in a network. The local availability of the functions and
their gradients necessitates coordination and consensus among the nodes. We put
forth the Generalized Gradient Tracking (GGT) framework that unifies a number
of existing approaches, including the state-of-the-art ones. The performance of
the proposed GGT algorithm is theoretically analyzed using a novel semidefinite
programming-based analysis that yields the desired regret bounds under very
general conditions and without requiring the gradient boundedness assumption.
The results are applicable to the special cases of GGT, which include various
state-of-the-art algorithms as well as new dynamic versions of various
classical decentralized algorithms. To further minimize the regret, we consider
a condensed version of GGT with only four free parameters. A procedure for
offline tuning of these parameters using only the problem parameters is also
detailed. The resulting optimized GGT (oGGT) algorithm not only achieves
improved dynamic regret bounds, but also outperforms all state-of-the-art
algorithms on both synthetic and real-world datasets.

本文提出了一种基于广义梯度跟踪（GGT）框架的去中心化在线学习算法，并使用新的半定编程分析理论对其性能进行了理论分析和优化，进而得到实际数据集上的优异表现。

分布式在线学习的优化梯度跟踪

Optimized Gradient Tracking for Decentralized Online Learning

In this paper we propose a novel framework for decentralized, online learning
by many learners. At each moment of time, an instance characterized by a
certain context may arrive to each learner; based on the context, the learner
can select one of its own actions (which gives a reward and provides
information) or request assistance from another learner. In the latter case,
the requester pays a cost and receives the reward but the provider learns the
information. In our framework, learners are modeled as cooperative contextual
bandits. Each learner seeks to maximize the expected reward from its arrivals,
which involves trading off the reward received from its own actions, the
information learned from its own actions, the reward received from the actions
requested of others and the cost paid for these actions - taking into account
what it has learned about the value of assistance from each other learner. We
develop distributed online learning algorithms and provide analytic bounds to
compare the efficiency of these with algorithms with the complete knowledge
(oracle) benchmark (in which the expected reward of every action in every
context is known by every learner). Our estimates show that regret - the loss
incurred by the algorithm - is sublinear in time. Our theoretical framework can
be used in many practical applications including Big Data mining, event
detection in surveillance sensor networks and distributed online recommendation
systems.

本文提出了一种新的分布式在线学习框架，将学习者建模为合作的情境赌博机，分析了分布式在线学习算法和完全知识基准的效率，研究表明后者在时间上失误是亚线性的，该理论框架可用于许多实际应用中，包括大数据挖掘、监视传感器网络事件检测和分布式在线推荐系统。