Point Process Methods Research Articles

Efficient Estimation of Directed Connectivity in Nonlinear and Nonstationary Spiking Neuron Networks.

Studying directed connectivity within spiking neuron networks can help understand neural mechanisms. Existing methods assume linear time-invariant neural dynamics with a fixed time lag in information transmission, while spiking networks usually involve complex dynamics that are nonlinear and nonstationary, and have varying time lags. We develop a Gated Recurrent Unit (GRU)-Point Process (PP) method to estimate directed connectivity within spiking networks. We use a GRU to describe the dependency of the target neuron's current firing rate on the source neurons' past spiking events and a PP to relate the target neuron's firing rate to its current 0-1 spiking event. The GRU model uses recurrent states and gate/activation functions to deal with varying time lags, nonlinearity, and nonstationarity in a parameter-efficient manner. We estimate the model using maximum likelihood and compute directed information as our measure of directed connectivity. We conduct simulations using artificial spiking networks and a biophysical model of Parkinson's disease to show that GRU-PP systematically addresses varying time lags, nonlinearity, and nonstationarity, and estimates directed connectivity with high accuracy and data efficiency. We also use a non-human-primate dataset to show that GRU-PP correctly identifies the biophysically-plausible stronger PMd-to-M1 connectivity than M1-to-PMd connectivity during reaching. In all experiments, the GRU-PP consistently outperforms state-of-the-art methods. The GRU-PP method efficiently estimates directed connectivity in varying time lag, nonlinear, and nonstationary spiking neuron networks. The proposed method can serve as a directed connectivity analysis tool for investigating complex spiking neuron network dynamics.

Spatial Heterogeneity of b Values in Northeastern Tibetan Plateau and Its Interpretation.

The northeastern margin of the Tibetan Plateau (NE Tibetan Plateau) exhibits active geological structures and has experienced multiple strong earthquakes, with M ≥ 7, throughout history. Particularly noteworthy is the 1920 M81/2 earthquake in the Haiyuan region that occurred a century ago and is documented as one of the deadliest earthquakes. Consequently, analyzing seismic risks in the northeastern margin of the Tibetan Plateau holds significant importance. The b value, a crucial parameter for seismic activity, plays a pivotal role in seismic hazard analyses. This study calculates the spatial b values in this region based on earthquake catalogs since 1970. The study area encompasses several major active faults, and due to variations in b values across different fault types, traditional grid-search methods may introduce significant errors in calculating the spatial b value within complex fault systems. To address this, we employed the hierarchical space-time point-process (HIST-PPM) method proposed by Ogata. This method avoids partitioning earthquake samples, optimizes parameters using Akaike's Bayesian Information Criterion (ABIC) with entropy maximization, and theoretically allows for a higher spatial resolution and more accurate b value calculations. The results indicate a high spatial heterogeneity in b values within the study area. The northwestern and southeastern regions exhibit higher b values. Along the Haiyuan fault zone, the central rupture zone of the Haiyuan earthquake has relatively higher b values than other regions of this fault zone, which is possibly related to the sufficient release of stress during the main rupture of the Haiyuan earthquake. The b values vary from high in the west to low in the east along the Zhongwei fault. On the West Qinling fault zone, the epicenter of the recent Minxian-Zhangxian earthquake is associated with a low b value. In general, regions with low b values correspond well to areas with moderate-strong seismic events in the past 50 years. The spatial differences in b values may reflect variances in seismic hazards among fault zones and regions within the same fault zone.

A novel point process model for neuronal spike trains

Point process provides a mathematical framework for characterizing neuronal spiking activities. Classical point process methods often focus on the conditional intensity function, which describes the likelihood at any time point given its spiking history. However, these models do not describe the central tendency or importance of the spike train observations. Based on the recent development on the notion of center-outward rank for point process, we propose a new modeling framework on spike train data. The new likelihood of a spike train is a product of the marginal probability on the number of spikes and the probability of spike timings conditioned on the same number. In particular, the conditioned distribution is calculated by adopting the well-known Isometric Log-Ratio transformation. We systematically compare the new likelihood with the state-of-the-art point process likelihoods in terms of ranking, outlier detection, and classification using simulations and real spike train data. This new framework can effectively identify templates as well as outliers in spike train data. It also provides a reasonable model, and the parameters can be efficiently estimated with conventional maximum likelihood methods. It is found that the proposed likelihood provides an appropriate ranking on the spike train observations, effectively detects outliers, and accurately conducts classification tasks in the given data.

Variational log-Gaussian point-process methods for grid cells.

We present practical solutions to applying Gaussian-process (GP) methods to calculate spatial statistics for grid cells in large environments. GPs are a data efficient approach to inferring neural tuning as a function of time, space, and other variables. We discuss how to design appropriate kernels for grid cells, and show that a variational Bayesian approach to log-Gaussian Poisson models can be calculated quickly. This class of models has closed-form expressions for the evidence lower-bound, and can be estimated rapidly for certain parameterizations of the posterior covariance. We provide an implementation that operates in a low-rank spatial frequency subspace for further acceleration, and demonstrate these methods on experimental data.

How Diverse Initial Samples Help and Hurt Bayesian Optimizers

Abstract Design researchers have struggled to produce quantitative predictions for exactly why and when diversity might help or hinder design search efforts. This paper addresses that problem by studying one ubiquitously used search strategy—Bayesian optimization (BO)—on a 2D test problem with modifiable convexity and difficulty. Specifically, we test how providing diverse versus non-diverse initial samples to BO affects its performance during search and introduce a fast ranked-determinantal point process method for computing diverse sets, which we need to detect sets of highly diverse or non-diverse initial samples. We initially found, to our surprise, that diversity did not appear to affect BO, neither helping nor hurting the optimizer’s convergence. However, follow-on experiments illuminated a key trade-off. Non-diverse initial samples hastened posterior convergence for the underlying model hyper-parameters—a model building advantage. In contrast, diverse initial samples accelerated exploring the function itself—a space exploration advantage. Both advantages help BO, but in different ways, and the initial sample diversity directly modulates how BO trades those advantages. Indeed, we show that fixing the BO hyper-parameters removes the model building advantage, causing diverse initial samples to always outperform models trained with non-diverse samples. These findings shed light on why, at least for BO-type optimizers, the use of diversity has mixed effects and cautions against the ubiquitous use of space-filling initializations in BO. To the extent that humans use explore-exploit search strategies similar to BO, our results provide a testable conjecture for why and when diversity may affect human-subject or design team experiments.

Multistream-Based Marked Point Process With Decomposed Cumulative Hazard Functions.

When applying a point process to a real-world problem, an appropriate intensity function model should be designed based on physical and mathematical prior knowledge. Recently, a fully trainable deep learning-based approach has been developed for temporal point processes. In this approach, a cumulative hazard function (CHF) capable of systematic computation of adaptive intensity function is modeled in a data-driven manner. However, in this approach, although many applications of point processes generate various kinds of information such as location, magnitude, and depth, the mark information of events is not considered. To overcome this limitation, we propose a fully trainable marked point process method for modeling decomposed CHFs for time and mark prediction using multistream deep neural networks. We demonstrate the effectiveness of the proposed method through experiments with synthetic and real-world event data.

FPGA-based point processing for denoising of the show-through effect in scanned document images

Document Image Binarization is one of the image enhancement techniques and has potential applications in the field of Document Image Analysis and Recognition (DIAR). During the DIAR process, different types of noise such as uneven illumination, clutter, margin, skew, bleed-through and show-through are removed. However, the removal of show-through noise is a challenging task because it affects mostly the document image quality and readability. In this work, we have designed and developed a model-based point processing operation, which is implemented in the Altera Cyclone® V DE1-SoC for removing the show-through noise and enhancing the degraded document image quality. The Altera Cyclone® V DE1-SoC board was chosen for this study due to its performance, size compatibility, and real-time implementation, which are better than the programmable DSP chips. The Pixel/Point processing operations were implemented using the DSP builder tool, which boost-up better execution of computational operations involved in complex algorithms. To implement the proposed algorithm, we have developed the board library file for Altera Cyclone® V DE1-SoC series 5CSEMA5F31C6 board in DSP builder and tested on various document images. The calculated quality measures of the proposed algorithm tested on various document images in the FPGA board exhibited better results in terms of document image readability, quality, processing speed, and memory size. Further, the performance of the proposed algorithms was studied in different scanned document images dataset as well as with the other traditional techniques. The results of this study witnessed that the proposed model-based point processing methods performs very well than the other methods.

Considering Neural Connectivity in Point Process Decoder for Brain-Machine Interface.

Brain machine interface (BMI) can translate neural activity into digital commands to control prostheses. The decoder in BMI models the mechanism relating to neural activity and intents in brain. In our brain, single neuronal tuning property and neural connectivity contribute to encoding the intents together. These properties may change, a phenomenon which is named neural adaptation during using BMIs. Neural adaptation requires the decoder to consider the two factors at the same time and has the potential to follow their changes. However, in the previous work, the class of neural network and clustering decoder can consider the neural connectivity regardless of the single neuronal tuning property. On the other hand, point process methods can model the single neuronal tuning property but fail to address the neural connectivity. In this paper, we propose a new point process decoder with the information of neural connectivity named NCPP. We derive the neural connectivity component from the point process method by Bayes' rule and use a clustering decoder to represent the neural connectivity. This method can consider the neural connectivity and the single neuronal tuning property at the same time. We validate the method on simulation data where the point process method cannot achieve a good decoding performance and compare it with sequential Monte Carlo point process method (SMCPP). The results show our method outperforms the pure point process method which indicates our method can model the neural connectivity and single neuronal tuning property at the same time.Clinical Relevance-This paper proposes a decoder that can model the neural connectivity and the single neuronal tuning property at the same time, which is potential to explain the neural adaptation computationally.

"Diabetic Retinopathy Early Detection Using Bag of Features Point Dependent Image Processing Methods"

Diabetic Retinopathy (DR) is a major problem that impairs anthropological eyesight. It is usually formed at the rear side of the eye retina, when lightsensitive tissues' blood vessels are damaged. Based on symptoms or diseases in the eye, such as mild visual difficulties, vision loss, blindness, damage to blood vessels in the retina caused by high blood sugar levels, etc., attention is required. Investigators have confronted a crucial test for the early identification of eye illness using image recognition methods for many years. A method for automatically identifying diabetic retinopathy based on topographies such as point-based from retinal pictures utilizing SVM and basket of topographies techniques. There are four steps in this procedure: The input eye retina picture is pre-processed in the first step. We smear rescaling geometric changes like revolution for picture enhancement in the second phase. In the third stage, we present an automated technique based on point-based features and a list of pictures for first diabetes or nondiabetic diagnosis utilizing SVM and a basket of topographies such as Haar modification image processing methods. Lastly, we liken the measurements like RMSE, PSNR, Elapsed Time and Accuracy utilizing this future method.

A methodology for alignment of measured rail profiles in turnouts as a basis for reliable vehicle/track interaction simulations

For the development of condition monitoring algorithms for turnouts, an extensive dataset consisting of measurement and simulation results is required. The basis for reliable simulation results are precise and easy applicable measurement techniques of the turnout geometry, like 3D handheld scanners. The challenge in using the data from portable 3D scanners for vehicle/track interaction simulation lies in the correct alignment and preparation of the measured surface. In this paper, a methodology for pre-processing of 3D scanner measurement results is presented, based on turnout measurements with the device HandySCAN 700 from Creaform. Point cloud processing methods like Principal Component Analysis and Iterative Closest Point algorithm are used to align the measured surface to a reference frame. Interpolation of the measured points by CAD reference data is applied for repairing the profile sections due to missing surface points. Furthermore, comparisons were made regarding the quality of simulation results for different input sources. It can be shown that the results have a good quality agreement in respect of different profile measurements techniques. Thus, it could be shown, that the data from the portable 3D scanner can be used for multibody dynamic simulation after appropriate preparation with the presented method.

An Algorithm for Large-Scale Multitarget Tracking and Parameter Estimation

Modern tracking problems require fast, scalable, and robust solutions for tracking multiple targets from noisy sensor data. In this article, an algorithm that has linear computational complexity with respect to the number of targets and measurements is presented. The method is based on the propagation of the first two factorial cumulants of a point process. The algorithm is demonstrated for tracking a million targets in cluttered environments in the fastest time yet for any such solution. A low-computational-complexity solution to the problem of joint multitarget tracking and parameter estimation is also presented. The multitarget filtering approach utilizes a single-cluster point process method for joint multiobject estimation and parameter estimation and is shown to be more computationally efficient and robust than previous implementations.

Tracking Fast Neural Adaptation by Globally Adaptive Point Process Estimation for Brain-Machine Interface.

Brain-machine interfaces (BMIs) help the disabled restore body functions by translating neural activity into digital commands to control external devices. Neural adaptation, where the brain signals change in response to external stimuli or movements, plays an important role in BMIs. When subjects purely use neural activity to brain-control a prosthesis, some neurons will actively explore a new tuning property to accomplish the movement task. The prediction of this neural tuning property can help subjects adapt more efficiently to brain control and maintain a good decoding performance. Existing prediction methods track the slow change of the tuning property in the manual control, which is not suitable for the fast neural adaptation in brain control. In order to identify the active neurons in brain control and track their tuning property changes, we propose a globally adaptive point process method (GaPP) to estimate the neural modulation state from spike trains, decompose the states into the hyper preferred direction and reconstruct the kinematics in a dual-model framework. We implement the method on real data from rats performing a two-lever discrimination task under manual control and brain control. The results show our method successfully predicts the neural modulation state and identifies the neurons that become active in brain control. Compared to the existing method, ours tracks the fast changes of the hyper preferred direction from manual control to brain control more accurately and efficiently and reconstructs the kinematics better and faster.

Properties of brightest group galaxies in cosmic web filaments

Context. The cosmic web, a complex network of galaxy groups and clusters connected by filaments, is a dynamical environment in which galaxies form and evolve. However, the impact of cosmic filaments on the properties of galaxies is difficult to study because of the much more influential local (galaxy-group scale) environment. Aims. The aim of this paper is to investigate the dependence of intrinsic galaxy properties on distance to the nearest cosmic web filament, using a sample of galaxies for which the local environment is easily assessable. Methods. Our study is based on a volume-limited galaxy sample with Mr ≤ −19 mag, drawn from the SDSS DR12. We chose brightest group galaxies (BGGs) in groups with two to six members as our probes of the impact of filamentary environment because their local environment can be determined more accurately. We use the Bisous marked point process method to detect cosmic-web filaments with radii of 0.5−1.0 Mpc and measure the perpendicular filament spine distance (Dfil) for the BGGs. We limit our study to Dfil values up to 4 Mpc. We use the luminosity density field as a tracer of the local environment. To achieve uniformity of the sample and to reduce potential biases we only consider filaments longer than 5 Mpc. Our final sample contains 1427 BGGs. Results. We note slight deviations between the galaxy populations inside and outside the filament radius in terms of stellar mass, colour, the 4000 Å break, specific star formation rates, and morphologies. However, all these differences remain below 95% confidence and are negligible compared to the effects arising from local environment density. Conclusions. Within a 4 Mpc radius of the filament axes, the effect of filaments on BGGs is marginal. The local environment is the main factor in determining BGG properties.

Application of 3D Laser Scanning Technology for Shield Tunnel Inspection

This paper describes the application of 3D laser scanning technology in shield tunnel inspection from the working mechanism. The point cloud data collection process, processing method and result analysis of the technology in shield tunnel inspection are introduced. Finally, combined with point cloud data obtained for dislocation and clearance inspection. The dislocation and clearance inspection are visually shown by the three - dimensional cloud chart. It shows that reliable inspection information can be obtained by using this technology, and has certain practical value for tunnel inspection.

Modeling Marked Temporal Point Process Using Multi-relation Structure RNN

Event sequences with marker and timing information are available in a wide range of domains, from machine log in automatic train supervision systems to information cascades in social networks. Given the historical event sequences, predicting what event will happen next and when it will happen can benefit many useful applications, such as maintenance service schedule for mass rapid transit trains and product advertising in social networks. Temporal point process (TPP) is one effective solution to solve the next event prediction problem due to its capability of capturing the temporal dependence among events. The recent recurrent temporal point process (RTPP) methods exploited recurrent neural network (RNN) to get rid of the parametric form assumption in the density functions of TPP. However, most existing RTPP methods focus only on the temporal dependence among events. In this work, we design a novel multi-relation structure RNN model with a hierarchical attention mechanism to capture not only the conventional temporal dependencies but also the explicit multi-relation topology dependencies. We then propose an RTPP algorithm whose density function conditioned on the event sequence embedding learned from our RNN model for cognitively predict the next event marker and time. The experiments show that our proposed MRS-RMTPP outperforms the state-of-the-art baselines in terms of both event marker prediction and event time prediction on three real-world datasets. The capability of capturing both ontology relation structure and temporal structure in the event sequences is of great importance for the next event marker and time prediction.

Spatial statistics for gaze patterns in scene viewing: Effects of repeated viewing

Scene viewing is used to study attentional selection in complex but still controlled environments. One of the main observations on eye movements during scene viewing is the inhomogeneous distribution of fixation locations: While some parts of an image are fixated by almost all observers and are inspected repeatedly by the same observer, other image parts remain unfixated by observers even after long exploration intervals. Here, we apply spatial point process methods to investigate the relationship between pairs of fixations. More precisely, we use the pair correlation function, a powerful statistical tool, to evaluate dependencies between fixation locations along individual scanpaths. We demonstrate that aggregation of fixation locations within 4° is stronger than expected from chance. Furthermore, the pair correlation function reveals stronger aggregation of fixations when the same image is presented a second time. We use simulations of a dynamical model to show that a narrower spatial attentional span may explain differences in pair correlations between the first and the second inspection of the same image.

Evaluation of mobile 3D light detection and ranging based canopy mapping system for tree fruit crops

In this study, 3D light detection and ranging (LiDAR) and an inertial measurement unit (IMU) were integrated on a ground vehicle for mapping tree fruit crops. A custom interface was developed in robot operating system for synchronous communication with hardware modules and for continuous field data collection. Point cloud data processing methods were developed for reconstruction and estimation of canopy parameters including height, voxel grid and convex hull methods-based volume and gap fractions in the canopy. The system was tested in apple tree and grapevine canopies. Overall, the mobile 3D LiDAR mapping system provided realistic representation of the canopy compared to the manual methods. For apple trees, the manual canopy volume measurements were strongly correlated to volume derived from the mobile 3D LiDAR mapping system (Convex hull: r = 0.81, Voxel grid: r = 0.51). The voxel grid method adequately considered gaps in the canopy during volume estimation and performed better than the convex hull method. The system was also evaluated for estimating canopy growth in grapevines with different rates of subsurface irrigation treatments. Deficit irrigation treatments did not show any significant effect on the canopy growth due to a high moisture content in the soil, resulting from high winter snowpack prior to that particular season. Nonetheless, the 3D LiDAR mapping system was able to aid in visualization of the temporal changes in canopy growth during the growing season. Change in vine canopy volume for the treatments followed a similar trend to the area ratio estimated from normalized differential vegetation index images derived from small unmanned aerial system based multispectral imagery. Overall, the 3D LiDAR based canopy mapping system and pertinent data mining algorithms can be the useful tool to the growers in rapid assessment of perennial fruit crop canopies for real-time management decision making.

Social-Aware Based Secure Relay Selection in Relay-Assisted D2D Communications

Relay-assisted transmission could effectively enhance the performance of Device-to-Device (D2D) communications when D2D user equipments (UEs) are too far away from each other or the quality of D2D channel is not good enough for direct communications. Meanwhile, security is one of the major concerns for proximity services. The secure relay selection problem for D2D communications under laying cellular network is studied in this paper. Firstly, we define a relay selection area and derive the closed-form of outage probability in D2D links using a Poisson Point Process (PPP) method. Next, in the defined relay selection area, we propose a secure relay selection scheme for the relay-assisted D2D communication system by exploiting the social relation as a security factor. Simulation results show that the scheme based on social relation can greatly improve the security performance of relay-assisted D2D communications.

JOINT PROCESSING OF ENVELOPE ALIGNMENT AND PHASE COMPENSATION FOR ISAR IMAGING

Abstract. Range envelope alignment and phase compensation are spilt into two isolated parts in the classical methods of translational motion compensation in Inverse Synthetic Aperture Radar (ISAR) imaging. In classic method of the rotating object imaging, the two reference points of the envelope alignment and the Phase Difference (PD) estimation are probably not the same point, making it difficult to uncouple the coupling term by conducting the correction of Migration Through Resolution Cell (MTRC). In this paper, an improved approach of joint processing which chooses certain scattering point as the sole reference point is proposed to perform with utilizing the Prominent Point Processing (PPP) method. With this end in view, we firstly get the initial image using classical methods from which a certain scattering point can be chose. The envelope alignment and phase compensation using the selected scattering point as the same reference point are subsequently conducted. The keystone transform is thus smoothly applied to further improve imaging quality. Both simulation experiments and real data processing are provided to demonstrate the performance of the proposed method compared with classical method.

Point Process Models for Quasi-Periodic Volcanic Earthquakes

Low frequency or long period (LP) earthquakes are a common phenomenon at active volcanoes, and are ubiquitous at persistently active andesitic and dacitic subduction zone volcanoes. At these systems, LP earthquakes provide critical information regarding the state of volcanic unrest, and their occurrence rates are key data on which eruption forecasts are based. Point process modeling of volcanic earthquake occurrence allows potential insights into the underlying physical processes driving unrest, and quantitative, probabilistic, eruption forecasts, for example, through application of the Failure Forecast Method (FFM). However, unlike high-frequency volcano-tectonic (VT) earthquakes, which are typically random or clustered in time, LPs are more commonly quasi-periodic or \textit{anti-clustered}. Consequently, the existing Poisson point process methods used to model occurrence rates of VT earthquakes are unlikely to be optimal for LP data. Here we evaluate the performance of candidate inhomogeneous point process formulations of the FFM for quasi-periodic LP data, based on four different inter-event time distributions: exponential (for Poisson), gamma, inverse Gaussian, and Weibull. Using example LP data recorded before a large explosion at Tungurahua volcano, Ecuador, we examine how well these models explain the observed data, and the quality of retrospective forecasts of eruption time. We use a Markov chain Monte Carlo approach to estimate parameter posterior distributions within a Bayesian framework. Goodness-of-fit is assessed using Quantile-Quantile and Kolmogorov-Smirnov methods, and results are benchmarked against those obtained from idealized synthetic datasets. Inverse Gaussian and gamma models were both found to fit the data well, with the inverse Gaussian model slightly outperforming the gamma model. However, retrospective forecasting analysis shows that the gamma model performs best, with the initial preference for the inverse Gaussian model controlled by catalog incompleteness late in the sequence. The gamma model fits the data significantly better than the Poisson model, and simulations show it produces better forecasts for highly periodic data. Simulations also show that forecast precision increases with the degree of periodicity of the earthquake process using the gamma model, and so should be better for LP earthquakes than VTs. These results provide a new framework for point process modeling of volcanic earthquake time series, and verification of alternative models.