Gradient Ascent Method Research Articles

Atrial Electrical Activity Detection Using Linear Combination of 12-Lead ECG Signals

ECG analysis is the method for cardiac arrhythmia diagnosis. During the diagnostic process many features should be taken into consideration, such as regularity and atrial activity. Since in some arrhythmias, the atrial electrical activity (AEA) waves are hidden in other waves, and a precise classification from surface ECG is inapplicable, a confirmation diagnosis is usually performed during an invasive procedure. In this paper, we study a "semiautomatic" method for AEA-waves detection using a linear combination of 12-lead ECG signals. This method's objective is to be applicable to a variety of arrhythmias with emphasis given to detect concealed AEA waves. It includes two variations--using maximum energy ratio and a synthetic AEA signal. In the former variation, an energy ratio-based cost function is created and maximized using the gradient ascent method. The latter variation adapted the linear combiner method, when applied on a synthetic signal, combined with surface ECG leads. A study was performed evaluating the AEA-waves detection from 63 patients (nine training, 54 validation) presenting eight arrhythmia types. Averaged sensitivity of 92.21% and averaged precision of 92.08% were achieved compared to the definite diagnosis. In conclusion, the presented method may lead to early and accurate detection of arrhythmias, which will result in a better oriented treatment.

Nearest Neighbor Classification with Locally Weighted Distance for Imbalanced Data

 Abstract—The datasets used in many real applications are highly imbalanced which makes classification problem hard. Classifying the minor class instances is difficult due to bias of the classifier output to the major classes. Nearest neighbor is one of the most popular and simplest classifiers with good performance on many datasets. However, correctly classifying the minor class is commonly sacrificed to achieve a better performance on others. This paper is aimed to improve the performance of nearest neighbor in imbalanced domains, without disrupting the real data distribution. Prototype-weighting is proposed, here, to locally adapting the distances to increase the chance of prototypes from minor class to be the nearest neighbor of a query instance. The objective function is, here, G-mean and optimization process is performed using gradient ascent method. Comparing the experimental results, our proposed method significantly outperformed similar works on 24 standard data sets.

Novel multimodality segmentation using level sets and Jensen‐Rényi divergence

Positron emission tomography (PET) is playing an increasing role in radiotherapy treatment planning. However, despite progress, robust algorithms for PET and multimodal image segmentation are still lacking, especially if the algorithm were extended to image-guided and adaptive radiotherapy (IGART). This work presents a novel multimodality segmentation algorithm using the Jensen-Rényi divergence (JRD) to evolve the geometric level set contour. The algorithm offers improved noise tolerance which is particularly applicable to segmentation of regions found in PET and cone-beam computed tomography. A steepest gradient ascent optimization method is used in conjunction with the JRD and a level set active contour to iteratively evolve a contour to partition an image based on statistical divergence of the intensity histograms. The algorithm is evaluated using PET scans of pharyngolaryngeal squamous cell carcinoma with the corresponding histological reference. The multimodality extension of the algorithm is evaluated using 22 PET/CT scans of patients with lung carcinoma and a physical phantom scanned under varying image quality conditions. The average concordance index (CI) of the JRD segmentation of the PET images was 0.56 with an average classification error of 65%. The segmentation of the lung carcinoma images had a maximum diameter relative error of 63%, 19.5%, and 14.8% when using CT, PET, and combined PET/CT images, respectively. The estimated maximal diameters of the gross tumor volume (GTV) showed a high correlation with the macroscopically determined maximal diameters, with a R(2) value of 0.85 and 0.88 using the PET and PET/CT images, respectively. Results from the physical phantom show that the JRD is more robust to image noise compared to mutual information and region growing. The JRD has shown improved noise tolerance compared to mutual information for the purpose of PET image segmentation. Presented is a flexible framework for multimodal image segmentation that can incorporate a large number of inputs efficiently for IGART.

A Faster Gradient Ascent Learning Algorithm for Nonlinear SVM

We propose a refined gradient ascent method including heuristic parameters for solving the dual problem of nonlinear SVM. Aiming to get better tuning to the particular training sequence, the proposed refinement consists of the use of heuristically established weights in correcting the search direction at each step of the learning algorithm that evolves in the feature space. We propose three variants for computing the correcting weights, their effectiveness being analyzed on experimental basis in the final part of the paper. The tests pointed out good convergence properties, and moreover, the proposed modified variants proved higher convergence rates as compared to Platt’s SMO algorithm. The experimental analysis aimed to derive conclusions on the recognition rate as well as on the generalization capacities. The learning phase of the SVM involved linearly separable samples randomly generated from Gaussian repartitions and the WINE and WDBC datasets. The generalization capacities in case of artificial data were evaluated by several tests performed on new linearly/nonlinearly separable data coming from the same classes. The tests pointed out high recognition rates (about 97%) on artificial datasets and even higher recognition rates in case of the WDBC dataset.

Use of measured scatter data for the attenuation correction of single photon emission tomography without transmission scanning

Attenuation correction is essential for reliable interpretation of emission tomography, however, the use of transmission measurements to generate attenuation maps is limited by availability of equipment and potential mismatches between the transmission and emission measurements. The authors present a first step toward a method of estimating an attenuation map from measured scatter data without a transmission scan. A scatter model has been developed that accurately predicts the distribution of photons which have been scattered once. The scatter model has been used as the basis of a maximum likelihood gradient ascent method to estimate an attenuation map from measured scatter data. In order to estimate both the attenuation map and activity distribution, iterations of the derived scatter based algorithm have been alternated with the maximum likelihood expectation maximization algorithm in a joint estimation process. For each iteration of the attenuation map estimation, the activity distribution is fixed at the values estimated during the previous activity iteration, and in each iteration of the activity distribution estimation the attenuation map is fixed at the values estimated during the previous attenuation iteration. The use of photopeak data to enhance the estimation of the attenuation map compared to the use of scatter data alone has also been considered. The algorithm derived has been used to reconstruct data simulated for an idealized two-dimensional situation and using a physical phantom. The reconstruction of idealized data demonstrated good reconstruction of both the activity distribution and attenuation map. The inclusion of information recorded in the photopeak energy window in the attenuation map estimation step demonstrated an improvement in the accuracy of the reconstruction, enabling an accurate attenuation map to be recovered. Validation of the results with physical phantom data demonstrated that different regions of attenuation could be distinguished in a real situation and produces results that represent a promising first step toward the use of scatter data to estimate the activity distribution and attenuation map from single photon emission tomography (SPECT) data without a transmission scan. The technique presented shows promise as a method of attenuation correction for SPECT data without the need for a separate transmission scan. Further work is required to further improve the method and to compensate for the assumptions used in the scatter model.

Principal Component Analysis by $L_{p}$ -Norm Maximization

This paper proposes several principal component analysis (PCA) methods based on Lp-norm optimization techniques. In doing so, the objective function is defined using the Lp-norm with an arbitrary p value, and the gradient of the objective function is computed on the basis of the fact that the number of training samples is finite. In the first part, an easier problem of extracting only one feature is dealt with. In this case, principal components are searched for either by a gradient ascent method or by a Lagrangian multiplier method. When more than one feature is needed, features can be extracted one by one greedily, based on the proposed method. Second, a more difficult problem is tackled that simultaneously extracts more than one feature. The proposed methods are shown to find a local optimal solution. In addition, they are easy to implement without significantly increasing computational complexity. Finally, the proposed methods are applied to several datasets with different values of p and their performances are compared with those of conventional PCA methods.

Integration of speckle de-noising and image segmentation using Synthetic Aperture Radar image for flood extent extraction

Flood is one of the detrimental hydro-meteorological threats to mankind. This compels very efficient flood assessment models. In this paper, we propose remote sensing based flood assessment using Synthetic Aperture Radar (SAR) image because of its imperviousness to unfavourable weather conditions. However, they suffer from the speckle noise. Hence, the processing of SAR image is applied in two stages: speckle removal filters and image segmentation methods for flood mapping. The speckle noise has been reduced with the help of Lee, Frost and Gamma MAP filters. A performance comparison of these speckle removal filters is presented. From the results obtained, we deduce that the Gamma MAP is reliable. The selected Gamma MAP filtered image is segmented using Gray Level Co-occurrence Matrix (GLCM) and Mean Shift Segmentation (MSS). The GLCM is a texture analysis method that separates the image pixels into water and non-water groups based on their spectral feature whereas MSS is a gradient ascent method, here segmentation is carried out using spectral and spatial information. As test case, Kosi river flood is considered in our study. From the segmentation result of both these methods are comprehensively analysed and concluded that the MSS is efficient for flood mapping.

Grid Task Scheduling Algorithm Based on Non-Conflict Degree

To solve the multi-objective conflict problem of independent task scheduling in grid computing, model the system along the direction of multi-objective gradient, give the model solution procedure, propose grid scheduling algorithm based on multi-objective non-conflict degree, that is, application of gradient ascent method does a quadratic optimization for the global optimal solution obtained by genetic algorithm, compensate for the GA (genetic algorithm) deficiency in the ability of local search. The experiments demonstrate that the proposed algorithm is better than other tasks scheduling algorithms simply using GA.

Fast numerical design of spatial-selective rf pulses in MRI using Krotov and quasi-Newton based optimal control methods

The use of increasingly strong magnetic fields in magnetic resonance imaging (MRI) improves sensitivity, susceptibility contrast, and spatial or spectral resolution for functional and localized spectroscopic imaging applications. However, along with these benefits come the challenges of increasing static field (B(0)) and rf field (B(1)) inhomogeneities induced by radial field susceptibility differences and poorer dielectric properties of objects in the scanner. Increasing fields also impose the need for rf irradiation at higher frequencies which may lead to elevated patient energy absorption, eventually posing a safety risk. These reasons have motivated the use of multidimensional rf pulses and parallel rf transmission, and their combination with tailoring of rf pulses for fast and low-power rf performance. For the latter application, analytical and approximate solutions are well-established in linear regimes, however, with increasing nonlinearities and constraints on the rf pulses, numerical iterative methods become attractive. Among such procedures, optimal control methods have recently demonstrated great potential. Here, we present a Krotov-based optimal control approach which as compared to earlier approaches provides very fast, monotonic convergence even without educated initial guesses. This is essential for in vivo MRI applications. The method is compared to a second-order gradient ascent method relying on the Broyden-Fletcher-Goldfarb-Shanno (BFGS) quasi-Newton method, and a hybrid scheme Krotov-BFGS is also introduced in this study. These optimal control approaches are demonstrated by the design of a 2D spatial selective rf pulse exciting the letters "JCP" in a water phantom.

Dimension Reduction by Mutual Information Feature Extraction

During the past decades, to study high-dimensional data in a large variety of problems, researchers have proposed many Feature Extraction algorithms. One of the most effective approaches for optimal feature extraction is based on mutual information (MI). However it is not always easy to get an accurate estimation for high dimensional MI. In terms of MI, the optimal feature extraction is creating a feature set from the data which jointly have the largest dependency on the target class and minimum redundancy. In this paper, a component-by-component gradient ascent method is proposed for feature extraction which is based on one-dimensional MI estimates. We will refer to this algorithm as Mutual Information Feature Extraction (MIFX). The performance of this proposed method is evaluated using UCI databases. The results indicate that MIFX provides a robust performance over different data sets which are almost always the best or comparable to the best ones .

Coverage Maximization with Autonomous Agents in Fast Flow Environments

This work examines the cooperative motion of a group of autonomous vehicles in a fast flow environment. The magnitude of the flow velocity is assumed to be greater than the available actuation to each agent. Collectively, the agents wish to maximize total coverage area defined as the set of points reachable by any agent within T time. The reachable set of an agent in a fast flow is characterized using optimal control techniques. Specifically, this work addresses the complementary cases where the static flow field is smooth, and where the flow field is piecewise constant. The latter case arises as a proposed approximation of a smooth flow that remains analytically tractable. Furthermore, the techniques used in the piecewise constant flow case enable treatment for obstacles in the environment. In both cases, a gradient ascent method is derived to maximize the total coverage area in a distributed fashion. Simulations show that such a network is able to maximize the coverage area in a fast flow.

A New Mean Shift Algorithm Based on Bacterial Colony Chemotaxis

Mean shift is an effective statistical iterative algorithm. Like other gradient ascent optimization methods, it is susceptible to local maxima, and hence often fails to find the desired global maximum. And in the iterative process, size of bandwidth has great impact on the accuracy and efficiency of the algorithm. It not only decides the number of sampling points in the iteration, but also affects the convergence speed and accuracy of the algorithm. For the above reason, a new mean shift algorithm based on bacterial colony chemotaxis (BCC) is proposed in this paper. Firstly, the mean shift vector is optimized using BCC algorithm. Then, the optimal mean shift vector is updated using mean shift procedure. For the choice of bandwidth, bandwidth is calculated by BCC algorithm. Experimental results show that the proposed algorithm used for image segmentation can segment images more effectively and provide more robust segmentation results.

Facial feature extraction using a probabilistic approach

Facial features such as lip corners, eye corners and nose tip are critical points in a human face. Robust extraction of such facial feature locations is an important problem which is used in a wide range of applications. In this work, we propose a probabilistic framework and several methods which can extract critical points on a face using both location and texture information. The new framework enables one to learn the facial feature locations probabilistically from training data. The principle is to maximize the joint distribution of location and apperance/texture parameters. We first introduce an independence assumption which enables independent search for each feature. Then, we improve upon this model by assuming dependence of location parameters but independence of texture parameters. We model location parameters with a multi-variate Gaussian and the texture parameters are modeled with a Gaussian mixture model which are much richer as compared to the standard subspace models like principal component analysis. The location parameters are found by solving a maximum likelihood optimization problem. We show that the optimization problem can be solved using various search strategies. We introduce local gradient-based methods such as gradient ascent and Newton's method initialized from independent model locations both of which require certain non-trivial assumptions to work. We also propose a multi-candidate coordinate ascent search and a coarse-to-fine search strategy which both depend on efficiently searching among multiple candidate points. Our framework is compared in detail with the conventional statistical approaches of active shape and active appearance models. We perform extensive experiments to show that the new methods outperform the conventional approaches in facial feature extraction accuracy.

Mean Shift Segmentation Algorithm Based on Hybridized Bacterial Chemotaxis

Mean shift, like other gradient ascent optimization methods, is susceptible to local maximum/minimum, and hence often fails to find the desired global maximum/minimum. For this reason, mean shift segmentation algorithm based on hybridized bacterial chemotaxis (HBC) is proposed in this paper. In HBC, particle swarm operation algorithm(PSO) is introduced before bacterial chemotaxis(BC) works. And PSO is firstly introduced to execute the global search, and then stochastic local search works by BC. Meanwhile, elitism preservation is used in the paper in order to improve the efficiency of the new algorithm. After mean shift vector is optimized using HBC algorithm, the optimal mean shift vector is updated using mean shift procedure. Experimental results show that new algorithm not only has higher convergence speed, but also can achieve more robust segmentation results.

A utility-based approach for secondary spectrum sharing

This paper provides a social welfare framework for coexistence of secondary users of spectrum in the presence of static primary users. We consider a formulation that captures spatial differences in available spectrum while considering general system topologies and utility functions: a collection of wireless sessions is considered under an arbitrary conflict graph that indicates the sessions which cannot transmit simultaneously on a common channel. It is assumed that each session has a utility associated with its spectrum utilization. A carrier sense multiple access-based randomized channel selection technique is considered to maximize the resulting sum of utilities. A measurement-based gradient ascent method is used to improve the channel selection performance and to achieve local maxima of the social welfare. Distributed versions of the method are discussed and shown to outperform previously published work in a variety of simulation scenarios that study effects of primary user presence, varying secondary user density, varying total channel availability.

Adaptive bandwidth mode detection algorithm

In this study a new algorithm ‘adaptive bandwidth mode detection’ (ABMD) algorithm has been developed to recover the correct density function without the need to either specify the correct number of Gaussians in the model or the correct bandwidth. The ABMD is employed in modelling visual features in applications such as image segmentation and real-time visual tracking. A simple type of model for these visual features are the Gaussian mixtures, where the number of Gaussian components is variable, thus, making it a flexible method for multimodal representation. This algorithm is used at initialisation for target modelling, where the target update will be done based on the mode propagation with adaptive bandwidth tracker method. It is based on an optimisation technique where a gradient ascent method is used and the optimal solution is selected based on a log-likelihood function. The mode detection ability of ABMD algorithm is compared with both the expectation maximisation and mean-shift algorithms. Furthermore, different video sequences have been employed to show how this approach has the ability to track an object regardless of whether the target model is corrupted with unwanted data at new frames.

Image registration and atlas-based segmentation of cardiac outflow velocity profiles

Cardiovascular disease is the leading cause of death worldwide and for this reason computer-based diagnosis of cardiac diseases is a very important task. In this article, a method for segmentation of aortic outflow velocity profiles from cardiac Doppler ultrasound images is presented. The proposed method is based on the statistical image atlas derived from ultrasound images of healthy volunteers. The ultrasound image segmentation is done by registration of the input image to the atlas, followed by a propagation of the segmentation result from the atlas onto the input image. In the registration process, the normalized mutual information is used as an image similarity measure, while optimization is performed using a multiresolution gradient ascent method. The registration method is evaluated using an in-silico phantom, real data from 30 volunteers, and an inverse consistency test. The segmentation method is evaluated using 59 images from healthy volunteers and 89 images from patients, and using cardiac parameters extracted from the segmented image. Experimental validation is conducted using a set of healthy volunteers and patients and has shown excellent results. Cardiac parameter segmentation evaluation showed that the variability of the automated segmentation relative to the manual is comparable to the intra-observer variability. The proposed method is useful for computer aided diagnosis and extraction of cardiac parameters.

Adaptive sun tracking algorithm for incident energy maximization and efficiency improvement of PV panels

In recent times, sun tracking systems are being increasingly employed to enhance the efficiency of photovoltaic panels by constantly tracking the elevation and azimuth angles of the sun. In this paper, a novel adaptive digital signal processing and control algorithm is presented that optimizes the overall PV system output power by adjusting the position angles of the solar panel on both the elevation and azimuth axes. Since the proposed approach is adaptive in nature, the optimal position angles for the solar panel are iteratively computed using the adaptive gradient ascent method, until the incident solar radiation, and hence the output power is maximized. Furthermore, a Taylor’s series approximation is employed for generating a unique optimal position angle increment/decrement at each iteration. Simulation results show that the proposed technique demonstrates fast convergence and excellent tracking accuracy at all times of the day.

Simulated visual homing in desert ant natural environments: efficiency of skyline cues

Desert ants, foraging in cluttered semiarid environments, are thought to be visually guided along individual, habitual routes. While other navigational mechanisms (e.g. path integration) are well studied, the question of how ants extract reliable visual features from a complex visual scene is still largely open. This paper explores the assumption that the upper outline of ground objects formed against the sky, i.e. the skyline, provides sufficient information for visual navigation. We constructed a virtual model of the ant's environment. In the virtual environment, panoramic images were recorded and adapted to the resolution of the desert ant's complex eye. From these images either a skyline code or a pixel-based intensity code were extracted. Further, two homing algorithms were implemented, a modified version of the average landmark vector (ALV) model (Lambrinos et al. Robot Auton Syst 30:39-64, 2000) and a gradient ascent method. Results show less spatial aliasing for skyline coding and best homing performance for ALV homing based on skyline codes. This supports the assumption of skyline coding in visual homing of desert ants and allows novel approaches to technical outdoor navigation.

An efficient method for large-scale slack allocation

This article concerns a timing or project graph, with given delays on the edges and given arrival times at the source and sink nodes. The arrival times at the other nodes are to be chosen; these determine the timing slacks, which must be non-negative, on the edges. The set of possible timing slacks is a polyhedron; to choose one, a separable concave utility function, such as the sum of the logarithms of the slacks, is maximized. This slack allocation problem, which can be given a simple statistical interpretation, is convex, and can be solved by a variety of methods. Gradient and coordinate ascent methods are simple and scale to large problems, but can converge slowly, depending on the topology and problem data. The Newton method, in contrast, reliably computes an accurate solution, but typically cannot scale beyond problems with a few thousand nodes. This article describes a custom truncated Newton method that efficiently computes an accurate solution, and scales to large graphs (say, with a million or more nodes). The method typically requires just a few hundred iterations, with each iteration requiring a few passes over the graph; in particular, the method has approximately linear complexity in the size of the problem. The same approach can be used to solve slack allocation problems with constraints, using an interior-point method that relies on the custom truncated Newton approach.