Search for: All records

Federated Learning (FL) revolutionizes collaborative machine learning among Internet of Things (IoT) devices by enabling them to train models collectively while preserving data privacy. FL algorithms fall into two primary categories: synchronous and asynchronous. While synchronous FL efficiently handles straggler devices, its convergence speed and model accuracy can be compromised. In contrast, asynchronous FL allows all devices to participate but incurs high communication overhead and potential model staleness. To overcome these limitations, the paper introduces a semi-synchronous FL framework that uses client tiering based on computing and communication latencies. Clients in different tiers upload their local models at distinct frequencies, striking a balance between straggler mitigation and communication costs. Building on this, the paper proposes the Dynamic client clustering, bandwidth allocation, and local training for semi-synchronous Federated learning (DecantFed) algorithm to dynamically optimize client clustering, bandwidth allocation, and local training workloads in order to maximize data sample processing rates in FL. DecantFed dynamically optimizes client clustering, bandwidth allocation, and local training workloads for maximizing data processing rates in FL. It also adapts client learning rates according to their tiers, thus addressing the model staleness issue. Extensive simulations using benchmark datasets like MNIST and CIFAR-10, under both IID and non-IID scenarios, demonstrate DecantFed’s superior performance. It outperforms FedAvg and FedProx in convergence speed and delivers at least a 28% improvement in model accuracy, compared to FedProx. 

Accurate quantum dynamics simulations of nonadiabatic processes are important for studies of electron transfer, energy transfer, and photochemical reactions in complex systems. In this comparative study, we benchmark various approximate nonadiabatic dynamics methods with mapping variables against numerically exact calculations based on the tensor-train (TT) representation of high-dimensional arrays, including TT-KSL for zero-temperature dynamics and TT-thermofield dynamics for finite-temperature dynamics. The approximate nonadiabatic dynamics methods investigated include mixed quantum–classical Ehrenfest mean-field and fewest-switches surface hopping, linearized semiclassical mapping dynamics, symmetrized quasiclassical dynamics, the spin-mapping method, and extended classical mapping models. Different model systems were evaluated, including the spin-boson model for nonadiabatic dynamics in the condensed phase, the linear vibronic coupling model for electronic transition through conical intersections, the photoisomerization model of retinal, and Tully’s one-dimensional scattering models. Our calculations show that the optimal choice of approximate dynamical method is system-specific, and the accuracy is sensitively dependent on the zero-point-energy parameter and the initial sampling strategy for the mapping variables.