Search results for: Gaussian distribution

Authors: Lakshmi Prayaga, Chandra Prayaga. Aaron Wade, Gopi Shankar Mallu, Harsha Satya Pola

Abstract:

Synthetic data generated by Generative Adversarial Networks and Autoencoders is becoming more common to combat the problem of insufficient data for research purposes. However, generating synthetic data is a tedious task requiring extensive mathematical and programming background. Open-source platforms such as the Synthetic Data Vault (SDV) and Mostly AI have offered a platform that is user-friendly and accessible to non-technical professionals to generate synthetic data to augment existing data for further analysis. The SDV also provides for additions to the generic GAN, such as the Gaussian copula. We present the results from two synthetic data sets (CTGAN data and CTGAN with Gaussian Copula) generated by the SDV and report the findings. The results indicate that the ROC and AUC curves for the data generated by adding the layer of Gaussian copula are much higher than the data generated by the CTGAN.

Keywords: synthetic data generation, generative adversarial networks, conditional tabular GAN, Gaussian copula

Procedia PDF Downloads 29

Authors: H. Hajimolahoseini, J. Hashemi, D. Redfearn

Abstract:

Background: Catheter ablation is an effective therapy for symptomatic atrial fibrillation (AF). The intracardiac electrocardiogram (IEGM) collected during this procedure contains precious information that has not been explored to its full capacity. Novel processing techniques allow looking at these recordings from different perspectives which can lead to improved therapeutic approaches. In our previous study, we showed that variation in amplitude measured through Shannon Entropy could be used as an AF recurrence risk stratification factor in patients who received Ibutilide before the electrograms were recorded. The aim of this study is to further investigate the effect of Ibutilide on characteristics of the recorded signals from the left atrium (LA) of a patient with persistent AF before and after administration of the drug. Methods: The IEGMs collected from different intra-atrial sites of 12 patients were studied and compared before and after Ibutilide administration. First, the before and after Ibutilide IEGMs that were recorded within a Euclidian distance of 3 mm in LA were selected as pairs for comparison. For every selected pair of IEGMs, the Probability Distribution Function (PDF) of the amplitude in time domain and magnitude in frequency domain was estimated using the regression analysis. The PDF represents the relative likelihood of a variable falling within a specific range of values. Results: Our observations showed that in time domain, the PDF of amplitudes was fitted to a Gaussian distribution while in frequency domain, it was fitted to a Rayleigh distribution. Our observations also revealed that after Ibutilide administration, the IEGMs would have significantly narrower short-tailed PDFs both in time and frequency domains. Conclusion: This study shows that the PDFs of the IEGMs before and after administration of Ibutilide represents significantly different properties, both in time and frequency domains. Hence, by fitting the PDF of IEGMs in time domain to a Gaussian distribution or in frequency domain to a Rayleigh distribution, the effect of Ibutilide can easily be tracked using the statistics of their PDF (e.g., standard deviation) while this is difficult through the waveform of IEGMs itself.

Keywords: atrial fibrillation, catheter ablation, probability distribution function, time-frequency characteristics

Procedia PDF Downloads 137

Authors: Reza Nasrollahpour, Mohamad Hidayat Bin Jamal, Zulhilmi Bin Ismail

Abstract:

Density currents are generated when the fluid of one density is released into another fluid with a different density. These currents occur in a variety of natural and man-made environments, and this emphasises the importance of studying them. In most practical cases, the density currents flow over the surfaces which are not plane; however, there have been limited investigations in this regard. This study uses laboratory experiments to analyse the influence of bottom roughness on the velocity distribution within these dense underflows. The currents are analysed over a plane surface and three different configurations of beam-roughened beds. The velocity profiles are collected using Acoustic Doppler Velocimetry technique, and the distribution of velocity within these currents is formulated for the tested beds. The results indicate that the empirical power and Gaussian relations can describe the velocity distribution in the inner and outer regions of the profiles, respectively. Moreover, it is found that the bottom roughness is the primary controlling parameter in the inner region.

Keywords: density currents, velocity profiles, Acoustic Doppler Velocimeter, bed roughness

Procedia PDF Downloads 150

Authors: Chandra Prayaga, Aaron Wade, Lakshmi Prayaga, Gopi Shankar Mallu

Abstract:

This paper presents the use of machine learning algorithms to predict the success of students in an introductory physics course. Data having 140 rows pertaining to the performance of two batches of students was used. The lack of sufficient data to train robust machine learning models was compensated for by generating synthetic data similar to the real data. CTGAN and CTGAN with Gaussian Copula (Gaussian) were used to generate synthetic data, with the real data as input. To check the similarity between the real data and each synthetic dataset, pair plots were made. The synthetic data was used to train machine learning models using the PyCaret package. For the CTGAN data, the Ada Boost Classifier (ADA) was found to be the ML model with the best fit, whereas the CTGAN with Gaussian Copula yielded Logistic Regression (LR) as the best model. Both models were then tested for accuracy with the real data. ROC-AUC analysis was performed for all the ten classes of the target variable (Grades A, A-, B+, B, B-, C+, C, C-, D, F). The ADA model with CTGAN data showed a mean AUC score of 0.4377, but the LR model with the Gaussian data showed a mean AUC score of 0.6149. ROC-AUC plots were obtained for each Grade value separately. The LR model with Gaussian data showed consistently better AUC scores compared to the ADA model with CTGAN data, except in two cases of the Grade value, C- and A-.

Keywords: machine learning, student success, physics course, grades, synthetic data, CTGAN, gaussian copula CTGAN

Procedia PDF Downloads 13

Authors: Septimia Sarbu

Abstract:

The entropy rate of a stochastic process is a fundamental concept in information theory. It provides a limit to the amount of information that can be transmitted reliably over a communication channel, as stated by Shannon's coding theorems. Recently, researchers have focused on developing new measures of information that generalize Shannon's classical theory. The aim is to design more efficient information encoding and transmission schemes. This paper continues the study of generalized entropy rates, by deriving a closed-form solution to the Sharma-Mittal entropy rate for Gaussian processes. Using the squeeze theorem, we solve the limit in the definition of the entropy rate, for different values of alpha and beta, which are the parameters of the Sharma-Mittal entropy. In the end, we compare it with Shannon and Rényi's entropy rates for Gaussian processes.

Keywords: generalized entropies, Sharma-Mittal entropy rate, Gaussian processes, eigenvalues of the covariance matrix, squeeze theorem

Procedia PDF Downloads 472

Authors: I. M. L. Nadeesha Jayaweera, Adao Alex Trindade

Abstract:

In choosing a candidate model in likelihood-based modeling via an information criterion, the practitioner is often faced with the difficult task of deciding just how far up the ranked list to look. Motivated by this pragmatic necessity, we construct an uncertainty band for a generalized (model selection) information criterion (GIC), defined as a criterion for which the limit in probability is identical to that of the normalized log-likelihood. This includes common special cases such as AIC & BIC. The method starts from the asymptotic normality of the GIC for the joint distribution of the candidate models in an independent and identically distributed (IID) data framework and proceeds by deriving the (asymptotically) exact distribution of the minimum. The calculation of an upper quantile for its distribution then involves the computation of multivariate Gaussian integrals, which is amenable to efficient implementation via the R package "mvtnorm". The performance of the methodology is tested on simulated data by checking the coverage probability of nominal upper quantiles and compared to the bootstrap. Both methods give coverages close to nominal for large samples, but the bootstrap is two orders of magnitude slower. The methodology is subsequently extended to two other commonly used model structures: regression and time series. In the regression case, we derive the corresponding asymptotically exact distribution of the minimum GIC invoking Lindeberg-Feller type conditions for triangular arrays and are thus able to similarly calculate upper quantiles for its distribution via multivariate Gaussian integration. The bootstrap once again provides a default competing procedure, and we find that similar comparison performance metrics hold as for the IID case. The time series case is complicated by far more intricate asymptotic regime for the joint distribution of the model GIC statistics. Under a Gaussian likelihood, the default in most packages, one needs to derive the limiting distribution of a normalized quadratic form for a realization from a stationary series. Under conditions on the process satisfied by ARMA models, a multivariate normal limit is once again achieved. The bootstrap can, however, be employed for its computation, whence we are once again in the multivariate Gaussian integration paradigm for upper quantile evaluation. Comparisons of this bootstrap-aided semi-exact method with the full-blown bootstrap once again reveal a similar performance but faster computation speeds. One of the most difficult problems in contemporary statistical methodological research is to be able to account for the extra variability introduced by model selection uncertainty, the so-called post-model selection inference (PMSI). We explore ways in which the GIC uncertainty band can be inverted to make inferences on the parameters. This is being attempted in the IID case by pivoting the CDF of the asymptotically exact distribution of the minimum GIC. For inference one parameter at a time and a small number of candidate models, this works well, whence the attained PMSI confidence intervals are wider than the MLE-based Wald, as expected.

Keywords: model selection inference, generalized information criteria, post model selection, Asymptotic Theory

Procedia PDF Downloads 57

Authors: Ahmad Alwosheel, Ahmed Alqaraawi

Abstract:

This work proposes an unsupervised text-independent probabilistic approach to recognize Quran reciter voice. It is an accurate approach that works on real time applications. This approach does not require a prior information about reciter models. It has two phases, where in the training phase the reciters' acoustical features are modeled using Gaussian Mixture Models, while in the testing phase, unlabeled reciter's acoustical features are examined among GMM models. Using this approach, a high accuracy results are achieved with efficient computation time process.

Keywords: Quran, speaker recognition, reciter recognition, Gaussian Mixture Model

Procedia PDF Downloads 349

Authors: Jyoti Wadhwa, Arvinder Singh

Abstract:

This work presents the theoretical investigation of an enhanced second-harmonic generation of higher-order Gaussian laser beam in plasma having a density ramp. The mechanism responsible for the self-focusing of a laser beam in plasma is considered to be the relativistic mass variation of plasma electrons under the effect of a highly intense laser beam. Using the moment theory approach and considering the Wentzel-Kramers-Brillouin approximation for the non-linear Schrodinger wave equation, the differential equation is derived, which governs the spot size of the higher-order Gaussian laser beam in plasma. The nonlinearity induced by the laser beam creates the density gradient in the background plasma electrons, which is responsible for the excitation of the electron plasma wave. The large amplitude electron plasma wave interacts with the fundamental beam, which further produces the coherent radiations with double the frequency of the incident beam. The analysis shows the important role of the different modes of higher-order Gaussian laser beam and density ramp on the efficiency of generated harmonics.

Keywords: density rippled plasma, higher order Gaussian laser beam, moment theory approach, second harmonic generation.

Procedia PDF Downloads 139

Authors: Prerana Sharma

Abstract:

Effect of relativistic nonlinearity on stimulated Raman scattering of the propagating laser beam carrying null intensity in center (hollow Gaussian beam) by excited plasma wave are studied in a collisionless plasma. The construction of the equations is done employing the fluid theory which is developed with partial differential equation and Maxwell’s equations. The analysis is done using eikonal method. The phenonmenon of Stimulated Raman scattering is shown along with the excitation of seed plasma wave. The power of plasma wave and back reflectivity is observed for higher order of hollow Gaussian beam. Back reflectivity is studied numerically for various orders of HGLB with different value of plasma density, laser power and beam radius. Numerical analysis shows that these parameters play vital role on reflectivity characteristics.

Keywords: Hollow Gaussian beam, relativistic nonlinearity, plasma physics, Raman scattering

Procedia PDF Downloads 602

Authors: Hacene Benkhoula, Mohamed Badreddine Benabdella, Hamid Bouzeboudja, Abderrahmane Asraoui

Abstract:

The wind is a random variable difficult to master, for this, we developed a mathematical and statistical methods enable to modeling and forecast wind power. Gaussian Processes (GP) is one of the most widely used families of stochastic processes for modeling dependent data observed over time, or space or time and space. GP is an underlying process formed by unrecognized operator’s uses to solve a problem. The purpose of this paper is to present how to forecast wind power by using the GP. The Gaussian process method for forecasting are presented. To validate the presented approach, a simulation under the MATLAB environment has been given.

Keywords: wind power, Gaussien process, modelling, forecasting

Procedia PDF Downloads 366

Authors: Wanatchapong Kongkaew

Abstract:

This paper proposes an application of probabilistic technique, namely Gaussian process regression, for estimating an optimal sequence of the single machine with total weighted tardiness (SMTWT) scheduling problem. In this work, the Gaussian process regression (GPR) model is utilized to predict an optimal sequence of the SMTWT problem, and its solution is improved by using an iterated local search based on simulated annealing scheme, called GPRISA algorithm. The results show that the proposed GPRISA method achieves a very good performance and a reasonable trade-off between solution quality and time consumption. Moreover, in the comparison of deviation from the best-known solution, the proposed mechanism noticeably outperforms the recently existing approaches.

Keywords: Gaussian process regression, iterated local search, simulated annealing, single machine total weighted tardiness

Procedia PDF Downloads 277

Authors: Mukesh Kumar Shah, Tushar Gupta

Abstract:

An Upgraded Cuckoo Search Algorithm is proposed here to solve optimization problems based on the improvements made in the earlier versions of Cuckoo Search Algorithm. Short comings of the earlier versions like slow convergence, trap in local optima improved in the proposed version by random initialization of solution by suggesting an Improved Lambda Iteration Relaxation method, Random Gaussian Distribution Walk to improve local search and further proposing Greedy Selection to accelerate to optimized solution quickly and by “Study Nearby Strategy” to improve global search performance by avoiding trapping to local optima. It is further proposed to generate better solution by Crossover Operation. The proposed strategy used in algorithm shows superiority in terms of high convergence speed over several classical algorithms. Three standard algorithms were tested on a 6-generator standard test system and the results are presented which clearly demonstrate its superiority over other established algorithms. The algorithm is also capable of handling higher unit systems.

Keywords: economic dispatch, gaussian selection operator, prohibited operating zones, ramp rate limits

Procedia PDF Downloads 100

Authors: Mohammadamin Abbasnejad

Abstract:

The analysis of expression and facial Action Units (AUs) detection are very important tasks in fields of computer vision and Human Computer Interaction (HCI) due to the wide range of applications in human life. Many works have been done during the past few years which has their own advantages and disadvantages. In this work, we present a new model based on Gaussian Conditional Random Field. We solve our objective problem using ADMM and we show how well the proposed model works. We train and test our work on two facial expression datasets, CK+, and RU-FACS. Experimental evaluation shows that our proposed approach outperform state of the art expression recognition.

Keywords: Gaussian Conditional Random Field, ADMM, convergence, gradient descent

Procedia PDF Downloads 319

Authors: Majid Habib Khan, Jinwei Zhao, Xinhong Hei, Liu Jiedong, Rana Shahzad Noor, Muhammad Imran

Abstract:

Deep learning is increasingly employed in image classification, yet the scarcity and high cost of labeled data for training remain a challenge. Limited samples often lead to overfitting due to biased sample distribution. This paper introduces a dynamic distribution calibration method for few-shot learning. Initially, base and new class samples undergo normalization to mitigate disparate feature magnitudes. A pre-trained model then extracts feature vectors from both classes. The method dynamically selects distribution characteristics from base classes (both adjacent and remote) in the embedding space, using a threshold value approach for new class samples. Given the propensity of similar classes to share feature distributions like mean and variance, this research assumes a Gaussian distribution for feature vectors. Subsequently, distributional features of new class samples are calibrated using a corrected hyperparameter, derived from the distribution features of both adjacent and distant base classes. This calibration augments the new class sample set. The technique demonstrates significant improvements, with up to 4% accuracy gains in few-shot classification challenges, as evidenced by tests on miniImagenet and CUB datasets.

Keywords: deep learning, computer vision, image classification, few-shot learning, threshold

Procedia PDF Downloads 25

Authors: Sunghoon Park, Hank Kim, Jinwon An, Sungzoon Cho

Abstract:

Internet of Things allows one to collect data from facilities which are then used to monitor them and even predict malfunctions in advance. Conventional quality control methods focus on setting a normal range on a sensor value defined between a lower control limit and an upper control limit, and declaring as an anomaly anything falling outside it. However, interactions among sensor values are ignored, thus leading to suboptimal performance. We propose a multivariate approach which takes into account many sensor values at the same time. In particular Gaussian Mixture Model is used which is trained to maximize likelihood value using Expectation-Maximization algorithm. The number of Gaussian component distributions is determined by Bayesian Information Criterion. The negative Log likelihood value is used as an anomaly score. The actual usage scenario goes like a following. For each instance of sensor values from a facility, an anomaly score is computed. If it is larger than a threshold, an alarm will go off and a human expert intervenes and checks the system. A real world data from Building energy system was used to test the model.

Keywords: facility anomaly detection, gaussian mixture model, anomaly score, expectation maximization algorithm

Procedia PDF Downloads 240

Authors: Sergio Curilef, Boris Atenas

Abstract:

Systems with long-range interactions are very common in nature. They are observed from the atomic scale to the astronomical scale and exhibit anomalies, such as inequivalence of ensembles, negative heat capacity, ergodicity breaking, nonequilibrium phase transitions, quasistationary states, and anomalous diffusion. These anomalies are exacerbated when special initial conditions are imposed; in particular, we use the so-called water bag initial conditions that stand for a uniform distribution. Several theoretical and practical implications are discussed here. A potential energy inspired by dipole-dipole interactions is proposed to build the dipole-type Hamiltonian mean-field model. As expected, the dynamics is novel and general to the behavior of systems with long-range interactions, which is obtained through molecular dynamics technique. Two plateaus sequentially emerge before arriving at equilibrium, which are corresponding to two different quasistationary states. The first plateau is a type of quasistationary state the lifetime of which depends on a power law of N and the second plateau seems to be a true quasistationary state as reported in the literature. The general behavior of the model according to its dynamics and thermodynamics is described. Using numerical simulation we characterize the mean kinetic energy, caloric curve, and the diffusion law through the mean square of displacement. The present challenge is to characterize the distributions in phase space. Certainly, the equilibrium state is well characterized by the Gaussian distribution, but quasistationary states in general depart from any Gaussian function.

Keywords: dipole-type interactions, dynamics and thermodynamics, mean field model, quasistationary states

Procedia PDF Downloads 177

Authors: Hossein Javidnia, Salehe Taheri

Abstract:

The Electrocardiogram (ECG) is the recording of the heart’s electrical potential versus time. ECG signals are often contaminated with noise such as baseline wander and muscle noise. As these signals have been widely used in clinical studies to detect heart diseases, it is essential to filter these noises. In this paper we compare performance of Wiener Filtering and Median Filtering methods to filter Additive White Gaussian (AWG) noise with the determined signal to noise ratio (SNR) ranging from 3 to 5 dB applied to long-term ECG recordings samples. Root mean square error (RMSE) and coefficient of determination (R2) between the filtered ECG and original ECG was used as the filter performance indicator. Experimental results show that Wiener filter has better noise filtering performance than Median filter.

Keywords: ECG noise filtering, Wiener filtering, median filtering, Gaussian noise, filtering performance

Procedia PDF Downloads 494

Authors: M. T. Alodat

Abstract:

In this paper, we propose a mechanism for skewing any symmetric distribution. The new distribution is called the deflation-inflation distribution (DID). We discuss some statistical properties of the DID such moments, stochastic representation, log-concavity. Also we fit the distribution to real data and we compare it to normal distribution and Azzlaini's skew normal distribution. Numerical results show that the DID fits the the tree ring data better than the other two distributions.

Keywords: normal distribution, moments, Fisher information, symmetric distributions

Procedia PDF Downloads 624

Authors: Ming Su, Ziqiang Mu

Abstract:

This paper proposes a synthesis method for nonuniform linear antenna arrays that combine Gaussian process regression (GPR) and genetic algorithm (GA). In this method, the GPR model can be used to calculate the array radiation pattern in the presence of mutual coupling effects, and then the GA is used to optimize the excitations and locations of the elements so as to generate the desired radiation pattern. In this paper, taking a 9-element nonuniform linear array as an example and the desired radiation pattern corresponding to a Chebyshev distribution as the optimization objective, optimize the excitations and locations of the elements. Finally, the optimization results are verified by electromagnetic simulation software CST, which shows that the method is effective.

Keywords: nonuniform linear antenna arrays, GPR, GA, mutual coupling effects, active element pattern

Procedia PDF Downloads 77

Authors: John Lorenzo Bautista, Yoon-Joong Kim

Abstract:

This paper presents the development of a Filipino speech recognition tool using the HTK System. The system was trained from a subset of the Filipino Speech Corpus developed by the DSP Laboratory of the University of the Philippines-Diliman. The speech corpus was both used in training and testing the system by estimating the parameters for phonetic HMM-based (Hidden-Markov Model) acoustic models. Experiments on different mixture-weights were incorporated in the study. The phoneme-level word-based recognition of a 5-state HMM resulted in an average accuracy rate of 80.13 for a single-Gaussian mixture model, 81.13 after implementing a phoneme-alignment, and 87.19 for the increased Gaussian-mixture weight model. The highest accuracy rate of 88.70% was obtained from a 5-state model with 6 Gaussian mixtures.

Keywords: Filipino language, Hidden Markov Model, HTK system, speech recognition

Procedia PDF Downloads 436

Authors: Kai Chen, Shuguang Cui, Feng Yin

Abstract:

Gaussian process (GP) with spectral mixture (SM) kernel demonstrates flexible non-parametric Bayesian learning ability in modeling unknown function. In this work a novel time-varying and non-stationary convolution spectral mixture (TN-CSM) kernel with a significant enhancing of interpretability by using process convolution is introduced. A way decomposing the SM component into an auto-convolution of base SM component and parameterizing it to be input dependent is outlined. Smoothly, performing a convolution between two base SM component yields a novel structure of non-stationary SM component with much better generalized expression and interpretation. The TN-CSM perfectly allows compatibility with the stationary SM kernel in terms of kernel form and spectral base ignored and confused by previous non-stationary kernels. On synthetic and real-world datatsets, experiments show the time-varying characteristics of hyper-parameters in TN-CSM and compare the learning performance of TN-CSM with popular and representative non-stationary GP.

Keywords: Gaussian process, spectral mixture, non-stationary, convolution

Procedia PDF Downloads 158

Authors: Soudabeh Shemehsavar

Abstract:

In this paper, we consider the situation under a life test, in which the failure time of the test units are not related deterministically to an observable stochastic time varying covariate. In such a case, the joint distribution of failure time and a marker value would be useful for modeling the step stress life test. The problem of accelerating such an experiment is considered as the main aim of this paper. We present a step stress accelerated model based on a bivariate Wiener process with one component as the latent (unobservable) degradation process, which determines the failure times and the other as a marker process, the degradation values of which are recorded at times of failure. Parametric inference based on the proposed model is discussed and the optimization procedure for obtaining the optimal time for changing the stress level is presented. The optimization criterion is to minimize the approximate variance of the maximum likelihood estimator of a percentile of the products’ lifetime distribution.

Keywords: bivariate normal, Fisher information matrix, inverse Gaussian distribution, Wiener process

Procedia PDF Downloads 291

Authors: Xiaoyu Shen, Fang Fang, Chujun Qiu

Abstract:

Herewith we present a Fourier method for credit risk quantification and allocation in the factor-copula model framework. The key insight is that, compared to directly computing the cumulative distribution function of the portfolio loss via Monte Carlo simulation, it is, in fact, more efficient to calculate the transformation of the distribution function in the Fourier domain instead and inverting back to the real domain can be done in just one step and semi-analytically, thanks to the popular COS method (with some adjustments). We also show that the Euler risk allocation problem can be solved in the same way since it can be transformed into the problem of evaluating a conditional cumulative distribution function. Once the conditional or unconditional cumulative distribution function is known, one can easily calculate various risk metrics. The proposed method not only fills the niche in literature, to the best of our knowledge, of accurate numerical methods for risk allocation but may also serve as a much faster alternative to the Monte Carlo simulation method for risk quantification in general. It can cope with various factor-copula model choices, which we demonstrate via examples of a two-factor Gaussian copula and a two-factor Gaussian-t hybrid copula. The fast error convergence is proved mathematically and then verified by numerical experiments, in which Value-at-Risk, Expected Shortfall, and conditional Expected Shortfall are taken as examples of commonly used risk metrics. The calculation speed and accuracy are tested to be significantly superior to the MC simulation for real-sized portfolios. The computational complexity is, by design, primarily driven by the number of factors instead of the number of obligors, as in the case of Monte Carlo simulation. The limitation of this method lies in the "curse of dimension" that is intrinsic to multi-dimensional numerical integration, which, however, can be relaxed with the help of dimension reduction techniques and/or parallel computing, as we will demonstrate in a separate paper. The potential application of this method has a wide range: from credit derivatives pricing to economic capital calculation of the banking book, default risk charge and incremental risk charge computation of the trading book, and even to other risk types than credit risk.

Keywords: credit portfolio, risk allocation, factor copula model, the COS method, Fourier method

Procedia PDF Downloads 116

Authors: Priya Kedia, Kiranmoy Das

Abstract:

There is a rich literature on variable selection in regression setting. However, most of these methods assume normality for the response variable under consideration for implementing the methodology and establishing the statistical properties of the estimates. In many real applications, the distribution for the response variable may be non-Gaussian, and one might be interested in finding the best subset of covariates at some predetermined quantile level. We develop dynamic Bayesian approach for variable selection in quantile regression framework. We use a zero-inflated mixture prior for the regression coefficients, and consider the asymmetric Laplace distribution for the response variable for modeling different quantiles of its distribution. An efficient Gibbs sampler is developed for our computation. Our proposed approach is assessed through extensive simulation studies, and real application of the proposed approach is also illustrated. We consider the data from health and retirement study conducted by the University of Michigan, and select the important predictors when the outcome of interest is out-of-pocket medical cost, which is considered as an important measure for financial risk. Our analysis finds important predictors at different quantiles of the outcome, and thus enhance our understanding on the effects of different predictors on the out-of-pocket medical cost.

Keywords: variable selection, quantile regression, Gibbs sampler, asymmetric Laplace distribution

Procedia PDF Downloads 120

Authors: Ravindra B. Patil, P. Krishnamoorthy, Shriram Sethuraman

Abstract:

This work proposes a novel Gaussian Mixture Model (GMM) based approach for accurate tracking of the arterial wall and subsequent computation of the distension waveform using Radio Frequency (RF) ultrasound signal. The approach was evaluated on ultrasound RF data acquired using a prototype ultrasound system from an artery mimicking flow phantom. The effectiveness of the proposed algorithm is demonstrated by comparing with existing wall tracking algorithms. The experimental results show that the proposed method provides 20% reduction in the error margin compared to the existing approaches in tracking the arterial wall movement. This approach coupled with ultrasound system can be used to estimate the arterial compliance parameters required for screening of cardiovascular related disorders.

Keywords: distension waveform, Gaussian Mixture Model, RF ultrasound, arterial wall movement

Procedia PDF Downloads 470

Authors: Donatella Giuliani

Abstract:

In this research, we propose an unsupervised grayscale image segmentation method based on a combination of the Firefly Algorithm and the Gaussian Mixture Model. Firstly, the Firefly Algorithm has been applied in a histogram-based research of cluster means. The Firefly Algorithm is a stochastic global optimization technique, centered on the flashing characteristics of fireflies. In this context it has been performed to determine the number of clusters and the related cluster means in a histogram-based segmentation approach. Successively these means are used in the initialization step for the parameter estimation of a Gaussian Mixture Model. The parametric probability density function of a Gaussian Mixture Model is represented as a weighted sum of Gaussian component densities, whose parameters are evaluated applying the iterative Expectation-Maximization technique. The coefficients of the linear super-position of Gaussians can be thought as prior probabilities of each component. Applying the Bayes rule, the posterior probabilities of the grayscale intensities have been evaluated, therefore their maxima are used to assign each pixel to the clusters, according to their gray-level values. The proposed approach appears fairly solid and reliable when applied even to complex grayscale images. The validation has been performed by using different standard measures, more precisely: the Root Mean Square Error (RMSE), the Structural Content (SC), the Normalized Correlation Coefficient (NK) and the Davies-Bouldin (DB) index. The achieved results have strongly confirmed the robustness of this gray scale segmentation method based on a metaheuristic algorithm. Another noteworthy advantage of this methodology is due to the use of maxima of responsibilities for the pixel assignment that implies a consistent reduction of the computational costs.

Keywords: clustering images, firefly algorithm, Gaussian mixture model, meta heuristic algorithm, image segmentation

Procedia PDF Downloads 188

Authors: N. Alhazmi

Abstract:

Knowing the thermodynamics properties of hydrocarbons is vital when it comes to analyzing the related chemical reaction outcomes and understanding the reaction process, especially in terms of petrochemical industrial applications, combustions, and catalytic reactions. However, measuring the thermodynamics properties experimentally is time-consuming and costly. In this paper, Gaussian process regression (GPR) has been used to directly predict the main thermodynamic properties - standard enthalpy of formation, standard entropy, and heat capacity -for more than 360 cyclic and non-cyclic alkanes, alkenes, and alkynes. A simple workflow has been proposed that can be applied to directly predict the main properties of any hydrocarbon by knowing its descriptors and chemical structure and can be generalized to predict the main properties of any material. The model was evaluated by calculating the statistical error R², which was more than 0.9794 for all the predicted properties.

Keywords: thermodynamic, Gaussian process regression, hydrocarbons, regression, supervised learning, entropy, enthalpy, heat capacity

Procedia PDF Downloads 186

Authors: R. J. Chang

Abstract:

A cyclostationary Gaussian linearization method is formulated for investigating the time average response of nonlinear system under sinusoidal signal and white noise excitation. The quantitative measure of cyclostationary mean, variance, spectrum of mean amplitude, and mean power spectral density of noise is analyzed. The qualitative response behavior of stochastic jump and bifurcation are investigated. The validity of the present approach in predicting the quantitative and qualitative statistical responses is supported by utilizing Monte Carlo simulations. The present analysis without imposing restrictive analytical conditions can be directly derived by solving non-linear algebraic equations. The analytical solution gives reliable quantitative and qualitative prediction of mean and noise response for the Duffing system subjected to both sinusoidal signal and white noise excitation.

Keywords: cyclostationary, duffing system, Gaussian linearization, sinusoidal, white noise

Procedia PDF Downloads 459

Authors: M. Mamor

Abstract:

GaN-based compounds have attracted much interest in the fabrication of high-power, high speed and high-frequency electronic devices. Other examples of GaN-based applications are blue and ultraviolet (UV) light-emitting diodes (LEDs). All these devices require high-quality ohmic and Schottky contacts. Gaining an understanding of the electrical characteristics of metal/GaN contacts is of fundamental and technological importance for developing GaN-based devices. In this work, the barrier characteristics of Pt and Pd Schottky contacts on n-type GaN were studied using temperature-dependent forward current-voltage (I-V) measurements over a wide temperature range 80–400 K. Our results show that the barrier height and ideality factor, extracted from the forward I-V characteristics based on thermionic emission (TE) model, exhibit an abnormal dependence with temperature; i.e., by increasing temperature, the barrier height increases whereas the ideality factor decreases. This abnormal behavior has been explained based on the TE model by considering the presence of double Gaussian distribution (GD) of nonhomogeneous barrier height at the metal/GaN interface. However, in the high-temperature range (160-400 K), the extracted value for the effective Richardson constant A* based on the barrier inhomogeneity (BHi) model is found in fair agreement with the theoretically predicted value of about 26.9 A.cm-2 K-2 for n-type GaN. This result indicates that in this temperature range, the conduction current transport is dominated by the thermionic emission mode. On the other hand, in the lower temperature range (80-160 K), the corresponding effective Richardson constant value according to the BHi model is lower than the theoretical value, suggesting the presence of other current transport, such as tunneling-assisted mode at lower temperatures.

Keywords: Schottky diodes, inhomogeneous barrier height, GaN semiconductors, Schottky barrier heights

Procedia PDF Downloads 21

Authors: Cheng Zeng, George Michailidis, Hitoshi Iyatomi, Leo L. Duan

Abstract:

The conditional density characterizes the distribution of a response variable y given other predictor x and plays a key role in many statistical tasks, including classification and outlier detection. Although there has been abundant work on the problem of Conditional Density Estimation (CDE) for a low-dimensional response in the presence of a high-dimensional predictor, little work has been done for a high-dimensional response such as images. The promising performance of normalizing flow (NF) neural networks in unconditional density estimation acts as a motivating starting point. In this work, the authors extend NF neural networks when external x is present. Specifically, they use the NF to parameterize a one-to-one transform between a high-dimensional y and a latent z that comprises two components [zₚ, zₙ]. The zₚ component is a low-dimensional subvector obtained from the posterior distribution of an elementary predictive model for x, such as logistic/linear regression. The zₙ component is a high-dimensional independent Gaussian vector, which explains the variations in y not or less related to x. Unlike existing CDE methods, the proposed approach coined Augmented Posterior CDE (AP-CDE) only requires a simple modification of the common normalizing flow framework while significantly improving the interpretation of the latent component since zₚ represents a supervised dimension reduction. In image analytics applications, AP-CDE shows good separation of 𝑥-related variations due to factors such as lighting condition and subject id from the other random variations. Further, the experiments show that an unconditional NF neural network based on an unsupervised model of z, such as a Gaussian mixture, fails to generate interpretable results.

Keywords: conditional density estimation, image generation, normalizing flow, supervised dimension reduction

Procedia PDF Downloads 60