No abstract

We analyse the convergence of the proximal gradient algorithm for convex composite problems in the presence of gradient and proximal computational inaccuracies. We derive new tighter deterministic and probabilistic bounds that we use to verify a simulated (MPC) and a synthetic (LASSO) optimization problems solved on a reduced-precision machine in combination with an inaccurate proximal operator. We also show how the probabilistic bounds are more robust for algorithm verification and more accurate for application performance guarantees. Under some statistical assumptions, we also prove that some cumulative error terms follow a martingale property. And conforming to observations, e.g., in [38], we also show how the acceleration of the algorithm amplifies the gradient and proximal computational errors.

No abstract

No abstract

We propose Dual-Feedback Generalized Proximal Gradient Descent (DFGPGD) as a new, hardware-friendly, operator splitting algorithm. We then establish convergence guarantees under approximate computational errors and we derive theoretical criteria for the numerical stability of DFGPGD based on absolute stability of dynamical systems. We also propose a new generalized proximal ADMM that can be used to instantiate most of existing proximal-based composite optimization solvers. We implement DFGPGD and ADMM on FPGA ZCU106 board and compare them in light of FPGA's timing as well as resource utilization and power efficiency. We also perform a full-stack, application-to-hardware, comparison between approximate versions of DFGPGD and ADMM based on dynamic power/error rate trade-off, which is a new hardware-application combined metric.

The objective of this work is to use a 1-dimensional signal that reflects the dissimilarity between multidimensional probability densities for detection. With the modified Kullback-Leibler divergence, faults can be directly detected without any normality assumption or joint monitoring of related test statistics in different subspaces such as the T 2 and SPE in principal component analysis-based methods. To relieve the difficulty associated with asymptotic high-dimensional density estimates, we have estimated the density ratio rather than the densities themselves. This can be done by approximating the density ratio with kernel basis functions and learn the weights from the available data. The developed algorithm is generic and can be applied to any industrial system as long as process historical data is available. As a case study, we apply this algorithm to a real rotary kiln in operation, which is an integral part of the cement manufacturing plant of Ain El Kebira, Algeria.

No abstract