Search

StochAstic Recursive grAdient algoritHm (SARAH), originally proposed for convex optimization and also proven to be effective for general nonconvex optimization, has received great attention because of its simple recursive framework for updating stochastic gradient estimates. The performance of SARAH significantly depends on the choice of the step size sequence. However, SARAH and its variants often manually select a best-tuned step size, which is time consuming in practice. Motivated by this gap, we propose a variant of the Barzilai-Borwein (BB) method, referred to as the Random Barzilai-Borwein (RBB) method, to determine the step size for SARAH in the mini-batch setting, leading to a new SARAH method: MB-SARAH-RBB. We prove that MB-SARAH-RBB converges linearly in expectation for strongly convex objective functions. Moreover, we analyze the gradient complexity of MB-SARAH-RBB and show that it is better than the original method. To further confirm the efficacy of the RBB method, we propose the MB-SARAH+-RBB method, by incorporating it into the MB-SARAH + method. Numerical experiments on standard data sets indicate that our proposed methods outperform or match state-of-the-art algorithms.

Pyrite and engineering carbon materials have received increasing attention for their catalytic potential in Fenton reactions due to their extensive sources and low cost. However, effects of carbon materials on the degradation of pollutants by pyrite-catalyzed heterogeneous Fenton oxidation have not been fully understood. In this study, the performance of pyrite-catalyzed heterogeneous Fenton system on the degradation of ciprofloxacin (CIP) was investigated in the presence of activated carbon (AC), biochar (BC), and carbon nanotubes (CNTs). Synchronous and asynchronous experiments (adsorption and catalysis) were conducted to elucidate the roles of the carbon materials in pyrite-catalyzed Fenton reactions. The results demonstrated that all the three carbon materials accelerated the pyrite-catalyzed Fenton oxidation of CIP. Under the experimental conditions, the reaction rates, which were obtained by fitting the synchronous experimental results with the pseudo-first-order kinetic model, of pyrite/AC, pyrite/BC and pyrite/CNTs with H

The ability of microplastics (MPs) to interact with environmental pollutants is of great concern. Riverine sediments, as sinks for multi-pollutants, have been rarely studied for MPs risk evaluation. Meanwhile, MPs generated from biodegradable plastics are questioning the safety of the promising materials. In this study, we investigated the effects of typical non-degradable polyethylene (PE) and biodegradable polylactic acid (PLA) MPs on sediment enzymes, arsenic (As) fractionation, and microbial community structures in As-contaminated riverine sediments. The results indicated that the presence of MPs (1% and 3%, w/w) led As transformed into more labile and bioavailable fractions in riverine sediments, especially under higher As and MPs levels. Analysis on microbial activities and community structures confirmed the strong potential of MPs in inhibiting microbial activities and shifting bacterial community succession patterns through enrichment of certain microbiota. Moreover, biodegradable PLA MPs presented stronger alterations in arsenic fractionation and microbial community structures than PE MPs did, which might be jointly attributed to adsorption behaviors, microbial alterations, and potential PLA degradation behaviors. The study indicated that MPs contamination increased As mobility and bioavailability, and shifted microbial communities in riverine sediments. Moreover, biodegradable MPs might lead to stronger microbial alterations and increases in As bioavailability, acting as a threat to ecological safety, which needed further exploration.

Plastic discharged into the environment would break down into microplastics (MPs). However, the possible impact of MPs on heavy metals in the aquatic sediment remains unknown. In order to evaluate the potential role of MPs as carriers of coexisting pollutants, the adsorption capacity of lead ions from sediment onto aged degradable and conventional MPs were investigated as a function of lead ions concentration, contact time, temperature, MPs dosage, aging time, and fulvic acid concentration. MPs were exposed to UV to obtain aged polyethylene (A-PE) and aged polylactic acid (A-PLA). The aging treatment increased the oxygen content, specific surface area and hydrophilicity of MPs. The adsorption capacity of A-PE for Pb(II) in sediment increased from 10.1525 to 10.4642 mg g−1 with the increasing aging time. However, the adsorption capacity of A-PLA for Pb(II) in sediment decreased from 9.3199 to 8.7231 mg g−1 with the increasing aging time. The adsorption capacity of MPs in sediment for Pb(II) was in the following order: A-PE > PLA > PE > A-PLA. Fulvic acid could promote the adsorption of Pb(II) by MPs in sediment. These results indicated that the aging process of the plastics in the environment would affect its role as a carrier of coexisting pollutants.

The treatment of dye pollutant and salts by nanofiltration membrane technology usually has the difficulty of low permeability and higher pressure in the case of maintaining a high rejection. In this study, high flux polyester membranes of β-cyclodextrin/trimesoyl chloride - multiwalled carbon nanotubes (β-CD/TMC-MWCNTs) were fabricated on a modified multiwalled carbon nanotubes (MWCNTs) substrate via interfacial polymerization (IP) process. The modified MWCNTs substrate was not only beneficial to modulating the permeation and distribution of aqueous phase monomer (β-cyclodextrin, β-CD), but also in favor of the formation of an ultrathin loose selective polyester active layer due to the porous network structure. The optimization experiments about the reaction time and monomer concentrations were conducted. The experimental results showed that the pure water flux of the as-prepared β-CD/TMC-MWCNTs membrane was 179.93 L m-2h-1 bar-1 (at the pressure of 0.1 MPa) when the reaction time of IP process was 5 min. The single dye removal efficiency of β-CD/TMC-MWCNTs membrane was 97.41% for Brilliant Green (BG) and 96.39% for Congo red (CR), respectively. The removal efficiency of dye/salt mixture by the membrane was above 84% for BG and over 98% for CR, respectively. The results indicated that the membrane had the efficient selective separation performance for the dye/salt mixture benefiting from the special molecular sieving architectures with the precisely intrinsic inner cavity of β-CD/TMC-MWCNTs membrane. Furthermore, the reusability, chlorine resistance, and antifouling ability of the β-CD/TMC-MWCNTs membrane was relatively excellent, and the membrane could effectively remove BG (>93%) and CR (>98%) simultaneously for mixed BG/CR dyes in real water body. These results suggested that the as-prepared β-CD/TMC-MWCNTs membrane had the potential of practical application in dye wastewater treatment.

Antimony (Sb) is the ubiquitous re-emerging contaminant greatly accumulated in sediments which has been revealed risky to ecological environment. However, the impacts of Sb (III/V) on microbes and plants in sediments, under different water management with presence of engineering materials are poorly understood. This study conducted sequential incubation of sediments (flooding, draining and planting) with presence of multiwall carbon nanotubes (MWCNTs) and Sb to explore the influence on microbial functional diversity, Sb accumulation and alfalfa traits. Results showed that water management and planting led to greater impacts of sediment enzyme activities and microbial community metabolic function and bioavailable Sb fractions (defined as sum of acid-soluble fraction and reducible fraction, F1 + F2). Available fractions of Sb (V) showed higher correlation to microbial metabolism (r = 0.933) than that of Sb (III) (r = −0.480) in planting stage. MWCNTs with increasing concentrations (0.01 1%, w/w) positively correlated to microbial community metabolic function in planting stage whereas resulted in decreasing of Sb (III/V) concentrations in alfalfa, although 0.01% MWCNT led to increase of Sb (V) and decrease of Sb (V) by 50.97% and 32.68% respectively. This study provided information for investigating combined ecological impacts of heavy metal and engineering materials under different water managing sediments.

In some multiźtarget tracking applications, appearing targets are suitably modeled as spawning from existing targets. However, in the original cardinalized probability hypothesis density (CPHD) filter, this type of model is not included; instead appearing targets are modeled by spontaneous birth only. Recently, two versions of CPHD filter modeling spawning targets have already been developed, but these two methods are tractable only when the spawning targets cardinality distribution is restricted to be the Bernoulli distribution, the Poisson distribution or the Zero-inflated Poisson distribution. In this paper, we derive a generalized CPHD filter which is tractable and has no constraint of the cardinality distribution of the spawning targets, that is to say, the spawning targets cardinality distribution can be arbitrary. The derivation is based on the finite set statistics (FISST) and the Faa di bruno's determinant formula. Moreover, how this generalized CPHD filter degrades into the two previous versions is also given in this paper. The resulting filter is different from the original CPHD filter in two aspects: first, the prediction equation of the PHD function changes to be identical with that of the probability hypothesis density (PHD) filter; and second, the cardinality distribution prediction equation is now an expression including the cardinality distribution information of the spawning targets. Simulation results show that the proposed method can response much faster than the original CPHD filter in target number estimate when spawning targets appear, and has a much smaller cardinality estimate variance than the PHD filter and the original CPHD filter. A comparison considering the optimal sub-pattern assignment (OSPA) metric also demonstrates the good performance of the proposed method. The general explicit cardinality prediction equation for the CPHD filter modeling spawning targets is derived.The derivation is based on the famous Faa di bruno's determinant formula.A tractable recursion computation technique of the general cardinality prediction equation is proposed.

Variance reduction techniques provide simple and fast algorithms for solving machine learning problems. In this paper, we present a novel stochastic variance-reduced method. The proposed method relies on the mini-batch version of stochastic recursive gradient algorithm (MB-SARAH), which updates stochastic gradient estimates by using a simple recursive scheme. However, facing the challenge of the step size sequence selection in MB-SARAH, we introduce an online step size sequence based on the hypergradient descent (HD) method, which only requires little additional computation. For the proposed method, referred to as MB-SARAH-HD, we provide a general convergence analysis and prove linear convergence for strongly convex problems in expectation. Specifically, we prove that the proposed method has sublinear convergence rate in a single outer loop. We also prove that the iteration complexity outperforms several variants of the state-of-the-art stochastic gradient descent (SGD) method under suitable conditions. Numerical experiments on standard datasets are provided to demonstrate the efficacy and superiority of our MB-SARAH-HD method over existing approaches in the literature.

In this paper, we focus on the reconfiguration of complex, intensive and sequential signal processing systems and propose a Multi-Stage Reconfigurable Signal Processing Flow (MSRSPF). The signal processing flow is partitioned into a number of stages which are connected with unified interfaces. With the combination of staged processing flow, the whole system can work at multiple modes with diverse functions and switch to a more suitable mode quickly according to the change of application requirements. As an application case, we introduce the usage of MSRSPF in passive signal reconnaissance. The reconnaissance system based on MSRSPF is flexible and adaptable to the electromagnetic environment, and can get the electronic information from different views with more reliability. The performance testing results show that MSRSPF supports multiple hardware logic and has a good reconfigurability with time overhead no more than a few seconds.

Mehrotra proposed a classical method for edge detection based on Gabor filter in 1992. He investigated the algorithm using the measures proposed by Canny and designed a criterion for one-dimensional (1-D) and two-dimensional (2-D) Gabor filter-based edge detector. The derivation was too complicated in time domain and it is not strict enough for 2-D step edge. We explain the Gabor filter-based edge detector by a more precise but much simpler mathematical derivation in frequency domain.