Differentially private SGD with random features

In the realm of large-scale machine learning,it is crucial to explore methods for reducing computational complexity and memory demands while maintaining generalization performance.Additionally,since the collected data may contain some sensitive information,it is also of great significance to study privacy-preserving machine learning algorithms.This paper focuses on the performance of the differentially private stochastic gradient descent(SGD)algorithm based on random features.To begin,the algorithm maps the original data into a lowdimensional space,thereby avoiding the traditional kernel method for large-scale data storage requirement.Subsequently,the algorithm iteratively optimizes parameters using the stochastic gradient descent approach.Lastly,the output perturbation mechanism is employed to introduce random noise,ensuring algorithmic privacy.We prove that the proposed algorithm satisfies the differential privacy while achieving fast convergence rates under some mild conditions.

Image Feature Extraction and Matching of Augmented Solar Images in Space Weather

Augmented solar images were used to research the adaptability of four representative image extraction and matching algorithms in space weather domain.These include the scale-invariant feature transform algorithm,speeded-up robust features algorithm,binary robust invariant scalable keypoints algorithm,and oriented fast and rotated brief algorithm.The performance of these algorithms was estimated in terms of matching accuracy,feature point richness,and running time.The experiment result showed that no algorithm achieved high accuracy while keeping low running time,and all algorithms are not suitable for image feature extraction and matching of augmented solar images.To solve this problem,an improved method was proposed by using two-frame matching to utilize the accuracy advantage of the scale-invariant feature transform algorithm and the speed advantage of the oriented fast and rotated brief algorithm.Furthermore,our method and the four representative algorithms were applied to augmented solar images.Our application experiments proved that our method achieved a similar high recognition rate to the scale-invariant feature transform algorithm which is significantly higher than other algorithms.Our method also obtained a similar low running time to the oriented fast and rotated brief algorithm,which is significantly lower than other algorithms.

Detecting soil salinity with arid fraction integrated index and salinity index in feature space using Landsat TM imagery

Modeling soil salinity in an arid salt-affected ecosystem is a difficult task when using remote sensing data because of the complicated soil context （vegetation cover, moisture, surface roughness, and organic matter） and the weak spectral features of salinized soil. Therefore, an index such as the salinity index （SI） that only uses soil spectra may not detect soil salinity effectively and quantitatively. The use of vegetation reflectance as an indirect indicator can avoid limitations associated with the direct use of soil reflectance. The normalized difference vegetation index （NDVI）, as the most common vegetation index, was found to be responsive to salinity but may not be available for retrieving sparse vegetation due to its sensitivity to background soil in arid areas. Therefore, the arid fraction integrated index （AFⅡ） was created as supported by the spectral mixture analysis （SMA）, which is more appropriate for analyzing variations in vegetation cover （particularly halophytes） than NDVI in the study area. Using soil and vegetation separately for detecting salinity perhaps is not feasible. Then, we developed a new and operational model, the soil salinity detecting model （SDM） that combines AFⅡ and SI to quantitatively estimate the salt content in the surface soil. SDMs, including SDM1 and SDM2, were constructed through analyzing the spatial characteristics of soils with different salinization degree by integrating AFⅡ and SI using a scatterplot. The SDMs were then compared to the combined spectral response index （COSRI） from field measurements with respect to the soil salt content. The results indicate that the SDM values are highly correlated with soil salinity, in contrast to the performance of COSRI. Strong exponential relationships were observed between soil salinity and SDMs （R2〉0.86, RMSE〈6.86） compared to COSRI （R2=0.71, RMSE=16.21）. These results suggest that the feature space related to biophysical properties combined with AFII and SI can effectively provide information on soil salinity.

Research of Underwater Bottom Object and Reverberation in Feature Space

The critical technical problem of underwater bottom object detection is founding a stable feature space for echo signals classification. The past literatures more focus on the characteristics of object echoes in feature space and reverberation is only treated as interference. In this paper, reverberation is considered as a kind of signal with steady characteristic, and the clustering of reverberation in frequency discrete wavelet transform （FDWT） feature space is studied. In order to extract the identifying information of echo signals, feature compression and cluster analysis are adopted in this paper, and the criterion of separability between object echoes and reverberation is given. The experimental data processing results show that reverberation has steady pattern in FDWT feature space which differs from that of object echoes. It is proven that there is separability between reverberation and object echoes.

Stacked spectral feature space patch: An advanced spectral representation for precise crop classification based on convolutional neural network

Spectral and spatial features in remotely sensed data play an irreplaceable role in classifying crop types for precision agriculture. Despite the thriving establishment of the handcrafted features, designing or selecting such features valid for specific crop types requires prior knowledge and thus remains an open challenge. Convolutional neural networks(CNNs) can effectively overcome this issue with their advanced ability to generate high-level features automatically but are still inadequate in mining spectral features compared to mining spatial features. This study proposed an enhanced spectral feature called Stacked Spectral Feature Space Patch(SSFSP) for CNN-based crop classification. SSFSP is a stack of twodimensional(2 D) gridded spectral feature images that record various crop types’ spatial and intensity distribution characteristics in a 2 D feature space consisting of two spectral bands. SSFSP can be input into2 D-CNNs to support the simultaneous mining of spectral and spatial features, as the spectral features are successfully converted to 2 D images that can be processed by CNN. We tested the performance of SSFSP by using it as the input to seven CNN models and one multilayer perceptron model for crop type classification compared to using conventional spectral features as input. Using high spatial resolution hyperspectral datasets at three sites, the comparative study demonstrated that SSFSP outperforms conventional spectral features regarding classification accuracy, robustness, and training efficiency. The theoretical analysis summarizes three reasons for its excellent performance. First, SSFSP mines the spectral interrelationship with feature generality, which reduces the required number of training samples.Second, the intra-class variance can be largely reduced by grid partitioning. Third, SSFSP is a highly sparse feature, which reduces the dependence on the CNN model structure and enables early and fast convergence in model training. In conclusion, SSFSP has great potential for practical crop classification in precision agriculture.

Feature Patch Illumination Spaces and Karcher Compression for Face Recognition via Grassmannians

Recent work has established that digital images of a human face, when collected with a fixed pose but under a variety of illumination conditions, possess discriminatory information that can be used in classification. In this paper we perform classification on Grassmannians to demonstrate that sufficient discriminatory information persists in feature patch (e.g., nose or eye patch) illumination spaces. We further employ the use of Karcher mean on the Grassmannians to demonstrate that this compressed representation can accelerate computations with relatively minor sacrifice on performance. The combination of these two ideas introduces a novel perspective in performing face recognition.

Discussion on the feature of strong earthquake: Orderly distribution in time, space and intensity before the Western Kunlun Mountain Pass M=8.1 earthquake

In the paper, the feature of strong earthquake orderly distribution in time, space and intensity before the Western Kunlun Mountain Pass M=8.1 earthquake is preliminarily studied. The modulation and triggering factors such as the earth rotation, earth tides are analyzed. The results show that: the giant earthquakes with the magnitude more than 8 occurred about every 24 years and the earthquakes with the magnitude more than 7 about every 7 years in Chinese mainland. The Western Kunlun Mountain M=8.1 earthquake exactly occurred at the expected time; The spatial distance show approximately the same distances between each two swarms. The earth rotation, earth tide, sun tide and sun magnetic field have played a role of modulation and triggering in the intensity. At last, the condi-tions for earthquake generation and occurrence are also discussed.

Reinforcement learning method for machining deformation control based on meta-invariant feature space

Precise control of machining deformation is crucial for improving the manufacturing quality of structural aerospace components.In the machining process,different batches of blanks have different residual stress distributions,which pose a significant challenge to machining deformation control.In this study,a reinforcement learning method for machining deformation control based on a meta-invariant feature space was developed.The proposed method uses a reinforcement-learning model to dynamically control the machining process by monitoring the deformation force.Moreover,combined with a meta-invariant feature space,the proposed method learns the internal relationship of the deformation control approaches under different stress distributions to achieve the machining deformation control of different batches of blanks.Finally,the experimental results show that the proposed method achieves better deformation control than the two existing benchmarking methods.

Feature mapping space and sample determination for person re-identification

Person re-identification(Re-ID) is integral to intelligent monitoring systems.However,due to the variability in viewing angles and illumination,it is easy to cause visual ambiguities,affecting the accuracy of person re-identification.An approach for person re-identification based on feature mapping space and sample determination is proposed.At first,a weight fusion model,including mean and maximum value of the horizontal occurrence in local features,is introduced into the mapping space to optimize local features.Then,the Gaussian distribution model with hierarchical mean and covariance of pixel features is introduced to enhance feature expression.Finally,considering the influence of the size of samples on metric learning performance,the appropriate metric learning is selected by sample determination method to further improve the performance of person re-identification.Experimental results on the VIPeR,PRID450 S and CUHK01 datasets demonstrate that the proposed method is better than the traditional methods.

A Comparative Study on Two Techniques of Reducing the Dimension of Text Feature Space

With the development of large scale text processing, the dimension of text feature space has become larger and larger, which has added a lot of difficulties to natural language processing. How to reduce the dimension has become a practical problem in the field. Here we present two clustering methods, i.e. concept association and concept abstract, to achieve the goal. The first refers to the keyword clustering based on the co occurrence of

Feature Extraction of Kernel Regress Reconstruction for Fault Diagnosis Based on Self-organizing Manifold Learning

The feature space extracted from vibration signals with various faults is often nonlinear and of high dimension.Currently,nonlinear dimensionality reduction methods are available for extracting low-dimensional embeddings,such as manifold learning.However,these methods are all based on manual intervention,which have some shortages in stability,and suppressing the disturbance noise.To extract features automatically,a manifold learning method with self-organization mapping is introduced for the first time.Under the non-uniform sample distribution reconstructed by the phase space,the expectation maximization(EM) iteration algorithm is used to divide the local neighborhoods adaptively without manual intervention.After that,the local tangent space alignment(LTSA) algorithm is adopted to compress the high-dimensional phase space into a more truthful low-dimensional representation.Finally,the signal is reconstructed by the kernel regression.Several typical states include the Lorenz system,engine fault with piston pin defect,and bearing fault with outer-race defect are analyzed.Compared with the LTSA and continuous wavelet transform,the results show that the background noise can be fully restrained and the entire periodic repetition of impact components is well separated and identified.A new way to automatically and precisely extract the impulsive components from mechanical signals is proposed.

Densification,microstructural features and tensile properties of selective laser melted AlMgSiScZr alloy from single track to block specimen

The selective laser melting(SLM)processed aluminum alloys have already aroused researchers’attention in aerospace,rail transport and petrochemical engineering due to the comprehensive advantages of low density,good corrosion resistance and high mechanical performance.In this paper,an Al-14.1Mg-0.47Si-0.31Sc-0.17Zr alloy was fabricated via SLM.The characteristics of single track at different process parameters,and the influence of hatch spacing on densification,microstructural features and tensile properties of block specimens were systematically studied.The hatch spacing has an influence on the overlap ratio of single track,and further affects the internal forming quality of printed specimen.At a laser power of 160 W and scanning speed of 400 mm/s,the densification of block specimen increased first and then decreased with the increase of hatch spacing.The nearly full dense specimen(98.7%)with a tensile strength of 452 MPa was fabricated at a hatch spacing of 80μm.Typical characteristics of dimple and cleavage on the tensile fracture of the AlMgSiScZr alloy showed the mixed fracture of ductility and brittleness.

Fractional Envelope Analysis for Rolling Element Bearing Weak Fault Feature Extraction

The bearing weak fault feature extraction is crucial to mechanical fault diagnosis and machine condition monitoring. Envelope analysis based on Hilbert transform has been widely used in bearing fault feature extraction. A generalization of the Hilbert transform, the fractional Hilbert transform is defined in the frequency domain, it is based upon the modification of spatial filter with a fractional parameter, and it can be used to construct a new kind of fractional analytic signal. By performing spectrum analysis on the fractional envelope signal, the fractional envelope spectrum can be obtained. When weak faults occur in a bearing, some of the characteristic frequencies will clearly appear in the fractional envelope spectrum. These characteristic frequencies can be used for bearing weak fault feature extraction. The effectiveness of the proposed method is verified through simulation signal and experiment data. © 2017 Chinese Association of Automation.

Multi-Scale Analysis Based Curve Feature Extraction in Reverse Engineering

A sectional curve feature extraction algorithm based on multi-scale analysis is proposed for reverse engineering. The algorithm consists of two parts: feature segmentation and feature classification. In the first part,curvature scale space is applied to multi-scale analysis and original feature detection. To obtain the primary and secondary curve primitives,feature fusion is realized by multi-scale feature detection information transmission. In the second part: projection height function is presented based on the area of quadrilateral,which improved criterions of sectional curve feature classification. Results of synthetic curves and practical scanned sectional curves are given to illustrate the efficiency of the proposed algorithm on feature extraction. The consistence between feature extraction based on multi-scale curvature analysis and curve primitives is verified.

Extraction of color-intensity feature towards image authentication

A color-intensity feature extraction method is proposed aimed at supplementing conventional image hashing algorithms that only consider intensity of the image. An image is mapped to a set of blocks represented by their dominant colors and average intensities. The dominant color is defined by hue and saturation with the hue value adjusted to make the principal colors more uniformly distributed. The average intensity is extracted from the Y component in the YCbCr space. By quantizing the color and intensity components, a feature vector is formed in a cylindrical coordinate system for each image block, which may be used to generate an intermediate hash. Euclidean distance is modified and a similarity metric introduced to measure the degree of similarity between images in terms of the color-intensity features. This is used to validate effectiveness of the proposed feature vector. Experiments show that the color-intensity feature is robust to normal image processing while sensitive to malicious alteration, in particular, color modification.

Study on the feature selection and classifica－tion effect of precursory data of ground tilt

Using the view point of nonlinear science and the method of selecting numerical features of pattern recognition for reference, the physical and numerical features of precursory ground tilt data are synthetically employed. The dynamic changes of data series are described with the numerical features in multi dimensional space and their distributive relations instead of an unique factor. The relationship between the ground tilt data and earthquake is examined through recognition and classification.

Case Study: Hydraulic Model Experiment to Analyze the Hydraulic Features for Installing Floating Islands

The viewpoint of a river is changing as people regard the river as water-friendly space where they can enjoy and share the space beyond the simple purpose of flood control alongside the improving social level. The floating islands installation was planned featuring three islands. The river’s flow and channel stability could be changed when new structures are built in a river. Hence an analysis of the hydraulic characteristic changes should need. The hydraulic model experiment in this study sought to review the impacts of the floating islands installation on the safety of flood control and stability of river channel. This study analyzed the hydraulic features affecting the surrounding stability when installing floating islands and proposed stable floating islands layout in terms of hydraulics based on the experiment results.

结合主动光源和改进YOLOv5s模型的夜间柑橘检测方法

【目的】解决夜间环境下遮挡和较小柑橘难以准确识别的问题,实现采摘机器人全天候智能化作业。【方法】提出一种结合主动光源的夜间柑橘识别方法。首先,通过分析主动光源下颜色特征不同的夜间柑橘图像,选择最佳的光源色并进行图像采集。然后,提出一种夜间柑橘检测模型BI-YOLOv5s,该模型采用双向特征金字塔网络(Bi-FPN)进行多尺度交叉连接和加权特征融合,提高对遮挡和较小果实的识别能力;引入Coordinate attention(CA)注意力机制模块,进一步加强对目标位置信息的提取;采用融入Transformer结构的C3TR模块,在减少计算量的同时更好地提取全局信息。【结果】本文提出的BI-YOLOv5s模型在测试集上的精准率、召回率、平均准确率分别为93.4%、92.2%和97.1%,相比YOLOv5s模型分别提升了3.2、1.5和2.3个百分点。在所采用的光源色环境下,模型对夜间柑橘识别的正确率为95.3%,相比白光环境下提高了10.4个百分点。【结论】本文提出的方法对夜间环境下遮挡和小目标柑橘的识别具有较高的准确性,可为夜间果蔬智能化采摘的视觉精准识别提供技术支持。

基于磁共振多模态成像定量参数鉴别诊断乳腺良恶性病变及与临床病理特征的关系探究

目的 分析磁共振多模态成像定量参数鉴别诊断乳腺良恶性病变的价值,并探究乳腺病变性质与临床病理特征的关系。方法 纳入2015年1月-2022年12月医院120例乳腺占位性病变患者,所有患者入院时均接受磁共振多模态成像检查,包括弥散加权(DWI)、磁共振波谱成像(MRS)与磁共振动态增强(DCE-MRI),以病理检查结果为“金标准”,分析该检查方式鉴别诊断乳腺良恶性病变的价值。依据乳腺占位性病变性质将患者分为恶性组与良性组,对比两组临床病理特征,分析乳腺病变性质与临床病理特征的关系。结果 120例乳腺占位性病变患者经病理检查结果显示,乳腺癌98例(81.67%),乳腺良性病变22例(18.33%);以病理检查结果为“金标准”,DCE-MRI及联合诊断乳腺病变性质的灵敏度、准确度高于DWI(P<0.05);乳腺癌组表观扩散系数(ADC)、rADC值低于乳腺良性病变组,Ⅰ型TIC曲线占比低于乳腺良性病变组,Ⅲ型TIC曲线占比高于乳腺良性病变组(P<0.05);绘制ROC曲线,结果显示,ADC、rADC及TIC曲线类型诊断乳腺病变性质具有一定价值(AUC=0.815、0.850、0.911);乳腺癌组肿瘤形态不规则、边缘毛刺/不规则、强化不均匀占比高于乳腺良性病变组(P<0.05);经Logistic回归分析,结果显示,肿瘤临床病理特征与乳腺占位性病变密切相关,形态不规则、边缘毛刺/不规则、强化不均匀是乳腺癌的危险因素(OR>1,P<0.05)。结论磁共振多模态成像定量参数对乳腺良恶性病变具有一定鉴别诊断价值,乳腺癌的发生与肿瘤形态不规则、边缘毛刺/不规则、强化不均匀等临床病理特征密切相关。