Entropy-based Technique Research Articles

Multistatic ISAR autofocus with an image entropy-based technique

A key issue in radar research for many years has been the noncooperative identification (NCI) of moving targets. Such targets can be aircraft, ships, ground vehicles, satellites, or ballistic missiles. Nowadays, identification is mostly based on cooperative methods like the identification friend-foe (IFF) systems in the military sector or automatic dependent surveillance-broadcast (ADS-B) for civil aircraft. However, a number of situations exist in which a cooperative identification is not possible. In conflict situations, hostile targets do not identify themselves. Knowledge of their type by NCI yields a tactical advantage. History has also shown a significant number of cases in which neutral or friendly targets have been destroyed due to their inability to identify themselves. Neutral targets usually do not share IFF codes of conflict parties, and friendly targets may turn off their IFF when performing covert operations. Operations between allied nations have also been a source of possible IFF failure. Furthermore, the desire for NCI is not limited to conflict situations. Unidentified aircraft or ships often are suspected to be terrorists. In all of the above situations, a decision on a possible response is extremely time critical, and possible consequences can be serious. For this reason, NCI remains an important research topic.

Market Efficiency, Roughness and Long Memory in PSI20 Index Returns: Wavelet and Entropy Analysis

In this study, features of the financial returns of the PSI20index, related to market efficiency, are captured using wavelet- and entropy-based techniques. This characterization includes the following points. First, the detection of long memory, associated with low frequencies, and a global measure of the time series: the Hurst exponent estimated by several methods, including wavelets. Second, the degree of roughness, or regularity variation, associated with the H¨older exponent, fractal dimension and estimation based on the multifractal spectrum. Finally, the degree of the unpredictability of the series, estimated by approximate entropy. These aspects may also be studied through the concepts of non-extensive entropy and distribution using, for instance, the Tsallis q-triplet. They allow one to study the existence of efficiency in the financial market. On the other hand, the study of local roughness is performed by considering wavelet leader-based entropy. In fact, the wavelet coefficients are computed from a multiresolution analysis, and the wavelet leaders are defined by the local suprema of these coefficients, near the point that we are considering. The resulting entropy is more accurate in that detection than the H¨older exponent. These procedures enhance the capacity to identify the occurrence of financial crashes.

BLACK-BOX TOOL FOR NONLINEAR SYSTEM IDENTIFICATION BASED UPON FUZZY SYSTEM

This paper introduces a novel identifier scheme for identification of nonlinear systems with disturbances. The identification process is carried out in two steps: an offline procedure and an online procedure. The method comprises of an automatic structure generating phase using entropybased technique. The accuracy of the model is suitably controlled using the entropy measure. The parameter learning phase uses the backpropagation technique. To improve the accuracy and also for generalization of the model to handle different data sets, Differential Evolution technique is employed whereby the parameters of the model are suitably tuned using evolutionary technique. A semi serial-parallel model is introduced to improve the online identification process in the presence of noisy data. The proposed mechanism is utilized and compared against the classical Sugeno, adaptive network-based fuzzy inference system (ANFIS) modeling and Laguerre Network-Based Fuzzy System for the identification of a nonlinear benchmark problem. In addition, the proposed technique is also used to model a rotary wing unmanned aerial vehicle (UAV) from real test input–output data. The modeling performance and generalization capability are seen to be superior with our method.

Estimating Regional Variations of R&D Effects on Productivity Growth by Entropy Econometrics

In this paper we adopt an entropy econometrics-based estimator to study regional variations in regression coefficients and apply it to analyse productivity growths generated by R&D activities at a regional level. Considering the possible effects of the region's own R&D stock as well as the spillovers produced in other regions, the paper proposes the use of an entropy-based technique to estimate these effects for a specific location. Depending on the degree of heterogeneity of the set of regions analysed, it is possible that some of these regions present characteristics that enable them to more easily convert R&D efforts (generated in the region itself or obtained from other regions by R&D spillovers) into productivity gains, whereas in other regions the effect of (direct or spillover generated) R&D activities may be irrelevant. We illustrate this idea with an empirical application for Spanish regions. RÉSUMÉ Dans la présente communication, nous adoptons un estimateur d'entropie à base économétrique pour l’étude de variations régionales des coefficients de régression et leur application pour l'analyse de d'augmentations de la productivité produites par des activités de recherche et développement à l’échelon régional. En se penchant sur les effets possibles du stock de recherches et développement propre à la région, ainsi que les retombées produites dans d'autres régions, la communication propose l'emploi d'une technique à base d'entropie pour effectuer une évaluation de ces effets pour un lieu spécifique. En fonction du degré d'hétérogénéité du groupe de régions analysé, il est possible que ces régions présentent des caractéristiques leur permettant de convertir plus facilement des interventions de recherche et développement (R&D) (générées dans cette même région, ou obtenues d'autres régions par retombées de R&D) en gains de productivité, alors que dans d'autres régions, l'effet (direct ou par retombée) d'activités de R&D pourra n'avoir aucune pertinence. Nous illustrons cette idée avec une application empirique pour des régions d'Espagne. EXTRACTO En este estudio adoptamos un estimador basado en máxima entropía para estudiar las variaciones regionales en coeficientes de regresión, y lo aplicamos al análisis de los crecimientos de productividad generados por actividades de I+D a nivel regional. Considerando los posibles efectos del stock de I+D propio de la región, así como los spillovers producidos en otras regiones, el estudio propone el uso de una técnica basada en entropía para estimar estos efectos en relación con una ubicación específica. Dependiendo del grado de heterogeneidad del conjunto de regiones analizadas, es posible que algunas de estas regiones presenten características que les permitan convertir más fácilmente los esfuerzos de I+D (generados en la región misma u obtenidos de otras regiones a través de spillovers de I+D) en ganancias de productividad, mientras que en otras regiones el efecto de las actividades de I+D (directas o generadas de spillovers) podrían ser irrelevantes. Ilustramos esta idea con una aplicación empírica para regiones españolas. 摘要: 本文中, 我们采用基于熵计量经济学的估计器研究回归系数的区域变化, 并用它分析地域级R&D活动带来的生产率增长。考虑到区域自身R&D储备和其他区域的人才流入可能带来影响, 本文提出一种熵方法估计这些因素对某个特定区域的影响。根据所分析区域集的异质程度, 其中有些区域可能会表现出一些特征, 使其更容易将R&D能力 (该区域自己产生或者从其他区域流入)转化为生产收益, 而另外一些区域R&D活动的影响 (直接或者流入)与生产收益无关。我们以对西班牙地区的实验应用为例证明了这一论点。

An entropy-LVQ system for S&P500 downward shifts forecasting

Article history: Received July 20, 2011 Accepted 7 October 2011 Available online 8 October 2011 The purpose of this paper is to predict the S&P500 down moves with technical analysis indicators using learning vector quantization (LVQ) neural networks and probabilistic neural networks (PNN). In addition, entropy-based input selection technique is employed to improve the prediction accuracies. The out-of-sample simulations show that LVQ outperforms PNN. In addition, the Entropy-LVQ system achieved higher accuracy in comparison with the literature. © 2012 Growing Science Ltd. All rights reserved.

Self-sustained vibrations in volcanic areas extracted by Independent Component Analysis: a review and new results

Abstract. We investigate the physical processes associated with volcanic tremor and explosions. A volcano is a complex system where a fluid source interacts with the solid edifice so generating seismic waves in a regime of low turbulence. Although the complex behavior escapes a simple universal description, the phases of activity generate stable (self-sustained) oscillations that can be described as a non-linear dynamical system of low dimensionality. So, the system requires to be investigated with non-linear methods able to individuate, decompose, and extract the main characteristics of the phenomenon. Independent Component Analysis (ICA), an entropy-based technique is a good candidate for this purpose. Here, we review the results of ICA applied to seismic signals acquired in some volcanic areas. We emphasize analogies and differences among the self-oscillations individuated in three cases: Stromboli (Italy), Erebus (Antarctica) and Volcán de Colima (Mexico). The waveforms of the extracted independent components are specific for each volcano, whereas the similarity can be ascribed to a very general common source mechanism involving the interaction between gas/magma flow and solid structures (the volcanic edifice). Indeed, chocking phenomena or inhomogeneities in the volcanic cavity can play the same role in generating self-oscillations as the languid and the reed do in musical instruments. The understanding of these background oscillations is relevant not only for explaining the volcanic source process and to make a forecast into the future, but sheds light on the physics of complex systems developing low turbulence.

A New Privacy Scheme for Providing Anonymity Techniques on Sensor Network

Research on sensor network had been focused on security services that provide authentication, confidentiality, integrity and availability until recently, but now there is growing interest in tackling the problem of actual sensor IDs being exposed. Many techniques for providing anonymity to the source in an ad-hoc network have been proposed, but they are not suitable. Thus, a technique that is well-suited to the characteristics of sensor networks is needed. This paper, limiting the type of attack against a sensor network to that of eavesdropping, proposes a new technique for providing anonymity using Phantom ID and SMAC. The degree of anonymity provided by the proposed technique was analyzed using an entropy-based modeling technique. The results showed that the anonymity is high when the proposed technique is used. The key factor responsible for the improved anonymity had to do with disguising the sensor ID so that it can’t be found easily.

Active entropy camera

Automatic exposure control (AEC) is traditionally based on the sampled intensity level in an image frame. However in certain situations, such intensity-based AEC does not achieve the desired results (with a maximum entropy) as shown in several examples in this paper. In most cases, maximizing entropy is the goal of a machine-vision system. We want the imaging system that captures the most amount of information, so that as many feature points as possible are presented in the captured image. To solve this problem, we propose an entropy-based exposure control technique. This technique adjusts the camera exposure setting to maximize the amount of information captured in the image frame. The amount of information is measured by Shannon entropy. The proposed entropy calculation is accelerated by graphics processing units (GPUs) to achieve a real-time performance. Experimental results show that under a number of conditions our AEC performs better than the traditional intensity-based AEC, preserving more detail, and capturing more information.

Entropy on covers

As a generalization of partitions, covers allow overlaps between their members. In this paper, we will propose a family of entropy-like measures over covers, which are anti monotonic with regard to the partial order defined by refinement relations of covers. In parallel to the entropy theory, we also develop their conditional forms, which in turn lead to a family of semi-metrics on covers. These make it possible to apply entropy-based techniques in data mining or machine learning to problems naturally modelled by covers.

Block-Entropy Analysis of Climate Data

We explore the use of block entropy as a dynamics classifier for meteorological timeseries data. The block entropy estimates define the entropy growth curve H(L) with respect to block length L. For a finitary process, the entropy growth curve tends to an asymptotic linear regime H(L) = E + hμL, with entropy rate hμ and excess entropy E. These quantities apportion the system's information content into ‘memory’ (E) and ‘randomness’ (hμ). We discuss the challenges inherent in analyzing weather data using symbolic techniques, identifying the pitfalls associated with alphabet size, finite sample timeseries length, and stationarity. We apply the block entropy-based techniques in the form of a wet/dry partition to Australian daily precipitation data from the Patched Point Dataset station record collection and version 3 of the Australian Water Availability Project analysis dataset. Preliminary results demonstrate hμ and E are viable climatological classifiers for precipitation, with station records from similar climatic regimes possessing similar values of hμ and E. The resultant clustering reflects expected characteristics of local climatic memory and randomness. The AWAP results show weaker clustering than their PPD counterparts, with different E- and hμ-values reflecting respectively the relative biases and truncation errors in the AWAP analysis system. The entropy rates of convergence analysis rules out finite order Markov processes for orders falling within the range of block sizes considered.

Density-Based Empirical Likelihood Ratio Change Point Detection Policies

Nonparametric detection of a change in distribution is studied when observations are independent. We provide a general method for constructing distribution-free change point detection schemes that have approximately likelihood structures. This method utilizes principles of the maximum empirical likelihood (EL) approach to approximate powerful parametric likelihood ratios. The product of the approximation can be associated with entropy-based test statistics. Entropy-based tests have been well addressed in the literature in the context of powerful decision rules for goodness of fit. We extend the entropy-based technique, using the EL principles. CUSUM and Shiryayev–Roberts (SR) detection policies are shown to be powerful parametric likelihood tests for detecting a change in distribution. We apply the proposed method to obtain nonparametric forms of the CUSUM and SR procedures. A Monte Carlo study demonstrates that the proposed method provides very efficient tests when comparing with the classical procedures. An example based on a real data is provided to demonstrate implementation and effectiveness of the new tests.

Adaptive mixture observation models for multiple object tracking

Multiple object tracking (MOT) poses many difficulties to conventional well-studied single object tracking (SOT) algorithms, such as severe expansion of configuration space, high complexity of motion conditions, and visual ambiguities among nearby targets, among which the visual ambiguity problem is the central challenge. In this paper, we address this problem by embedding adaptive mixture observation models (AMOM) into a mixture tracker which is implemented in Particle Filter framework. In AMOM, the extracted multiple features for appearance description are combined according to their discriminative power between ambiguity prone objects, where the discriminability of features are evaluated by online entropy-based feature selection techniques. The induction of AMOM can help to surmount the incapability of conventional mixture tracker in handling object occlusions, and meanwhile retain its merits of flexibility and high efficiency. The final experiments show significant improvement in MOT scenarios compared with other methods.

Internet Traffic Behavior Profiling for Network Security Monitoring

Recent spates of cyber-attacks and frequent emergence of applications affecting Internet traffic dynamics have made it imperative to develop effective techniques that can extract, and make sense of, significant communication patterns from Internet traffic data for use in network operations and security management. In this paper, we present a general methodology for building comprehensive behavior profiles of Internet backbone traffic in terms of communication patterns of end-hosts and services. Relying on data mining and entropy-based techniques, the methodology consists of significant cluster extraction, automatic behavior classification and structural modeling for in-depth interpretive analyses. We validate the methodology using data sets from the core of the Internet.

Uniform rotational motion compensation for inverse synthetic aperture radar with non-cooperative targets

In the current scenario of high-range resolution radars, the rotational motion of the target usually generates migration through resolution cells (MTRC) in the inverse synthetic aperture radar (ISAR) images. A technique to correct the MTRC in the case of uniform rotation rate is proposed. An estimation of the rotation vector of the non-cooperative targets is not required. The approach compensates the rotational motion in two steps: the slant-range rotation compensation (SRRC) and the cross-range rotation compensation (CRRC). CRRC is based on an extension of phase difference (PD) and on a robust method to delete outliers. The technique is verified both with simulated and live data from a millimetre-wave linear frequency-modulated continuous wave radar. A comparison with a minimum entropy-based technique is also made.

Very-Long-Period Volcanic Tremor at Stromboli, Italy

We analyze a long time–space series of Stromboli volcanic tremors. A very-low-frequency content in the range of 0.02–0.5 Hz has been found by using spectral analysis and independent component analysis. Independent component analysis is an entropy-based technique. We observe the occurrence of a component having a period of 30 sec. Polarization analysis shows that the wave field comes mainly from the crater area, well evidenced by seismometers located around the summit ring, whereas the radiation becomes increasingly scattered at stations located around the base of this volcano. Based on its apparent velocity, the 30-sec component appears to be a slow wave, related to inhomogeneities of the source and/or gas-pressure fluctuations inside the shallow plumbing system.

Evaluation of Entropy-based Image Co-registration Techniques for Brain Perfusion SPECT

We examined image co-registration techniques based on mutual information (MI) and normalized mutual information (NMI) to registrate two SPECT images with localized differences in brain activity. Simulations were performed by 100 trials with the random initial mismatch of ±20° and ±45 mm for a normal perfusion model (NRM) and a pathological perfusion model with a hemisphere defect (HSD) . In the HSD, activity in the defect region was reduced to be 75% (HSD75%), 50% (HSD50%), and 25% (HSD25%) of normal value. In the MI-based registration for the NRM and NRM dataset, no biases were observed (≤-0.06°, ≤0.06 mm) and the SDs were very small (≤0.07°, ≤0.04 mm) . In the registration for the NRM and HSD dataset, the biases were slightly larger (≤-0.18°, ≤0.34 mm for HSD75%, ≤-0.23°, ≤=0.45 mm for HSD50%, ≤-0.23°, ≤0.18 mm for HSD25%) in comparison with the results of registration for the NRM and NRM dataset. The accuracy of the NMI was almost identical to that of the MI. The entropy-based registration techniques are relatively unaffected by localized differences in brain activity.

Evolutionary spectrum estimation by positivity constrained deconvolution

We present a deconvolution technique to obtain the evolutionary spectrum (ES) of nonstationary signals by deconvolving the blurring effects of the time-frequency distribution (TFD) kernel from bilinear TFDs. The resulting spectrum is non-negative and has desirable properties such as higher resolution and higher concentration in time frequency. The new technique is computationally more efficient compared with the previously proposed entropy-based deconvolution technique, and, unlike the entropy method, it is not restricted to deconvolution of spectrograms with Gaussian windows. This makes the method applicable to deconvolving many of the bilinear time-frequency distributions.

Entropy thresholding and its parallel algorithm on the reconfigurable array of processors with wider bus networks

Thresholding is the most commonly used technique in image segmentation. We first propose an efficient sequential algorithm to improve the relative entropy-based thresholding technique. This algorithm combines the concepts of the relative entropy with that of the local entropy and also includes the quadtree hierarchical structure in it. Second, we derive a constant time parallel algorithm to solve this problem on the reconfigurable array of processors with wider bus networks (RAPWBN). The system bus bandwidth determines the capacity of data communication between processors. According to the results as shown by Li and Maresca (1989) and by Maresca and Li (1989), we know that the silicon area used by the switching control mechanism is far less than that used by the processor. Instead of increasing the number of processors, we extend the number of buses to increase the power of a parallel processing system. Such a strategy of utilizing the reconfigurable array of processors with wider bus networks not only has the advantage of saving silicon area but also increases the system power enormously. So, we use the RAPWBN to solve the entropy-based thresholding problem.