面向第三方支付平台非结构化大数据分布特征的融合聚类算法

对于蓬勃发展的第三方支付平台,挖掘非结构化数据中蕴含的分布信息有助于平台对商户精准细分。为充分挖掘多元交易分布信息,本文提出针对多元分布样本的商户聚类算法。本文首先采用高斯混合模型拟合交易分布,进而基于Wasserstein距离提出多元分布间的距离定义,并设计迭代算法实现商户聚类。在此基础上,本文进一步提出融合多元分布与协变量特征的聚类算法,得到更综合稳健的聚类效果。本文分别在模拟数据和某第三方支付平台真实商户交易数据上进行验证,结果表明本文算法相较于其他对比方法具有更准确的聚类表现,并且通过聚类结果还能从分布、协变量特征等角度直观反映异常套现商户的行为模式,为第三方支付平台进行风险识别、个性化管理等提供决策支持。此方法可广泛应用于需要融合多维度数据分析的客群聚类场景。

Tuning Interface Bridging Between MoSe2 and Three‑Dimensional Carbon Framework by Incorporation of MoC Intermediate to Boost Lithium Storage Capability

Interface engineering has been widely explored to improve the electrochemical performances of composite electrodes,which governs the interface charge transfer,electron transportation,and structural stability.Herein,MoC is incorporated into MoSe2/C composite as an intermediate phase to alter the bridging between MoSe2-and nitrogen-doped three-dimensional(3D)carbon framework as MoSe2/MoC/N–C connection,which greatly improve the structural stability,electronic conductivity,and interfacial charge transfer.Moreover,the incorporation of MoC into the composites inhibits the overgrowth of MoSe2 nanosheets on the 3D carbon framework,producing much smaller MoSe2 nanodots.The obtained MoSe2 nanodots with fewer layers,rich edge sites,and heteroatom doping ensure the good kinetics to promote pseudo-capacitance contributions.Employing as anode material for lithium-ion batteries,it shows ultralong cycle life(with 90%capacity retention after 5000 cycles at 2 A g−1)and excellent rate capability.Moreover,the constructed LiFePO4//MoSe2/MoC/N–C full cell exhibits over 86%capacity retention at 2 A g−1 after 300 cycles.The results demonstrate the effectiveness of the interface engineering by incorporation of MoC as interface bridging intermediate to boost the lithium storage capability,which can be extended as a potential general strategy for the interface engineering of composite materials.

Recurrent neural network-based volumetric fluorescence microscopy

Volumetric imaging of samples using fluorescence microscopy plays an important role in various fields including physical,medical and life sciences.Here we report a deep learning-based volumetric image inference framework that uses 2D images that are sparsely captured by a standard wide-field fluorescence microscope at arbitrary axial positions within the sample volume.Through a recurrent convolutional neural network,which we term as Recurrent-MZ,2D fluorescence information from a few axial planes within the sample is explicitly incorporated to digitally reconstruct the sample volume over an extended depth-of-field.Using experiments on C.elegans and nanobead samples,Recurrent-MZ is demonstrated to significantly increase the depth-of-field of a 63×/1.4NA objective lens,also providing a 30-fold reduction in the number of axial scans required to image the same sample volume.We further illustrated the generalization of this recurrent network for 3D imaging by showing its resilience to varying imaging conditions,including e.g.,different sequences of input images,covering various axial permutations and unknown axial positioning errors.We also demonstrated wide-field to confocal cross-modality image transformations using Recurrent-MZ framework and performed 3D image reconstruction of a sample using a few wide-field 2D fluorescence images as input,matching confocal microscopy images of the same sample volume.Recurrent-MZ demonstrates the first application of recurrent neural networks in microscopic image reconstruction and provides a flexible and rapid volumetric imaging framework,overcoming the limitations of current 3D scanning microscopy tools.

Bright-field holography:cross-modality deep learning enables snapshot 3D imaging with bright-field contrast using a single hologram

Digital holographic microscopy enables the 3D reconstruction of volumetric samples from a single-snapshot hologram.However,unlike a conventional bright-field microscopy image,the quality of holographic reconstructions is compromised by interference fringes as a result of twin images and out-of-plane objects.Here,we demonstrate that cross-modality deep learning using a generative adversarial network(GAN)can endow holographic images of a sample volume with bright-field microscopy contrast,combining the volumetric imaging capability of holography with the speckle-and artifact-free image contrast of incoherent bright-field microscopy.We illustrate the performance of this“bright-field holography”method through the snapshot imaging of bioaerosols distributed in 3D,matching the artifact-free image contrast and axial sectioning performance of a high-NA bright-field microscope.This data-driven deep-learning-based imaging method bridges the contrast gap between coherent and incoherent imaging,and enables the snapshot 3D imaging of objects with bright-field contrast from a single hologram,benefiting from the wave-propagation framework of holography.

Motility-based label-free detection of parasites in bodily fluids using holographic speckle analysis and deep learning

Parasitic infections constitute a major global public health issue.Existing screening methods that are based on manual microscopic examination often struggle to provide sufficient volumetric throughput and sensitivity to facilitate early diagnosis.Here,we demonstrate a motility-based label-free computational imaging platform to rapidly detect motile parasites in optically dense bodily fluids by utilizing the locomotion of the parasites as a specific biomarker and endogenous contrast mechanism.Based on this principle,a cost-effective and mobile instrument,which rapidly screens~3.2 mL of fluid sample in three dimensions,was built to automatically detect and count motile microorganisms using their holographic time-lapse speckle patterns.We demonstrate the capabilities of our platform by detecting trypanosomes,which are motile protozoan parasites,with various species that cause deadly diseases affecting millions of people worldwide.Using a holographic speckle analysis algorithm combined with deep learningbased classification,we demonstrate sensitive and label-free detection of trypanosomes within spiked whole blood and artificial cerebrospinal fluid(CSF)samples,achieving a limit of detection of ten trypanosomes per mL of whole blood(~five-fold better than the current state-of-the-art parasitological method)and three trypanosomes per mL of CSF.We further demonstrate that this platform can be applied to detect other motile parasites by imaging Trichomonas vaginalis,the causative agent of trichomoniasis,which affects 275 million people worldwide.With its costeffective,portable design and rapid screening time,this unique platform has the potential to be applied for sensitive and timely diagnosis of neglected tropical diseases caused by motile parasites and other parasitic infections in resource-limited regions.

Neural network-based image reconstruction in swept-source optical coherence tomography using undersampled spectral data

Optical coherence tomography(OCT)is a widely used non-invasive biomedical imaging modality that can rapidly provide volumetric images of samples.Here,we present a deep learning-based image reconstruction framework that can generate swept-source OCT(SS-OCT)images using undersampled spectral data,without any spatial aliasing artifacts.This neural network-based image reconstruction does not require any hardware changes to the optical setup and can be easily integrated with existing swept-source or spectral-domain OCT systems to reduce the amount of raw spectral data to be acquired.To show the efficacy of this framework,we trained and blindly tested a deep neural network using mouse embryo samples imaged by an SS-OCT system.Using 2-fold undersampled spectral data(i.e.,640 spectral points per A-line),the trained neural network can blindly reconstruct 512 A-lines in 0.59 ms using multiple graphics-processing units(GPUs),removing spatial aliasing artifacts due to spectral undersampling,also presenting a very good match to the images of the same samples,reconstructed using the full spectral OCT data(i.e.,1280 spectral points per A-line).We also successfully demonstrate that this framework can be further extended to process 3×undersampled spectral data per A-line,with some performance degradation in the reconstructed image quality compared to 2×spectral undersampling.Furthermore,an A-line-optimized undersampling method is presented by jointly optimizing the spectral sampling locations and the corresponding image reconstruction network,which improved the overall imaging performance using less spectral data points per A-line compared to 2×or 3×spectral undersampling results.This deep learning-enabled image reconstruction approach can be broadly used in various forms of spectral-domain OCT systems,helping to increase their imaging speed without sacrificing image resolution and signal-to-noise ratio.