Search NASA⌕ Search

SEARCH · Search NASA

Results for “Log-Likelihood Function”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

15 records
All sourcesNASA NTRSDOE OSTI
NASA NTRS1980

Improving the performance of soft decision Viterbi decoding in a non-Gaussian environment through non-linear quantization

The performance of Viterbi decoding in a non-Gaussian environment is investigated using a nonlinear quantization strategy. The channel model consists of a convolutionally encoded BPSK signal transmitted to a satellite where it is corrupted with additive white Gaussian noise and pulsed radio frequency interference (RFI). The resultant signal is then passed through a satellite nonlinearity and transmitted to a ground station where it is coherently detected. Interleaving is assumed in order to make the channel memoryless. The presence of RFI makes the channel statistics non-Gaussian, leading to a nonlinear log-likelihood function. A near optimum quantization scheme is found by maximizing a channel parameter, or by matching the quantizer to the log-likelihood function in a mean square error sense. Bit error rate performance improvement is achieved by using such nonlinear quantization.

Mcgregor, D. N.↗
DOE OSTI2022

Transfer Learning-Based Independent Component Analysis

Understanding the underlying component structure is crucial for multivariate signal analysis. Among all the techniques that try to learn the latent structure, independent component analysis (ICA) is one of the most important and popular methods, which aims to extract independent components from multivariate signals and enables further analysis. For example, in electroencephalogram (EEG) analysis, artifacts filtering and disease detection are conducted based on the independent components of the signals. One critical challenge in existing ICA approaches is that the component extraction accuracy may degrade when the available data of a unit are limited. To address this issue, this paper proposes a transfer learning-based ICA method by innovatively transferring component distribution from a source domain, so that accurate component extraction results can be achieved even when only limited data are available in the target domain. To the best of our knowledge, this is the first work that leverages transfer learning to improve ICA accuracy with limited available data. In particular, we first extract all the independent components from the source domain by maximizing the log-likelihood function with a Newton-like method on a smooth manifold. Then for the target domain, the component with the largest negentropy is extracted in each round. To effectively leverage the knowledge from the source domain and to prevent the negative transfer, we try to find a component in the source domain that matches the component we are extracting. The probability density function of the matched component will then be used to improve the component extraction accuracy if such matched component can be found; otherwise, no knowledge will be transferred. Finally, numerical simulations and a case study with electrocardiogram (ECG) data are conducted, showing the effectiveness of the proposed method in transferring knowledge and reducing negative transfer.

42 ENGINEERING↗
DOE OSTI2024

Maximum Likelihood Estimation: Some Basics

The maximum likelihood estimation is a general estimation procedure. It is often compared to estimation procedures like the ordinary least squares regression or generalized method of moments, to name a few. We discuss some basics about the maximum likelihood estimation, its advantages and disadvantages, and provide an example application to a gamma distribution function.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2025

Maximum Likelihood Estimation: Some Basics

The maximum likelihood estimation is a general estimation procedure. It is often compared to estimation procedures like the ordinary least squares regression or generalized method of moments, to name a few. We discuss some basics about the maximum likelihood estimation, its advantages and disadvantages, and provide an example application to a gamma distribution function.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1974

A square root formulation for the combined state-parameter estimator with application to the identification of sailplane performance

A square root formulation is presented for the discrete combined state parameter estimation problem with linear plant dynamics, Gaussian random disturbances, and constant but uncertain parameters. The estimator is a combination of the classical Kalman filter and a maximum likelihood algorithm which maximizes the parameter log-likelihood function using a first order search routine.

Froidevaux, M. R.↗
DOE OSTI2021

Training Restricted Boltzmann Machines With a D-Wave Quantum Annealer

Restricted Boltzmann Machine (RBM) is an energy-based, undirected graphical model. It is commonly used for unsupervised and supervised machine learning. Typically, RBM is trained using contrastive divergence (CD). However, training with CD is slow and does not estimate the exact gradient of the log-likelihood cost function. In this work, the model expectation of gradient learning for RBM has been calculated using a quantum annealer (D-Wave 2000Q), where obtaining samples is faster than Markov chain Monte Carlo (MCMC) used in CD. Training and classification results of RBM trained using quantum annealing are compared with the CD-based method. The performance of the two approaches is compared with respect to the classification accuracies, image reconstruction, and log-likelihood results. The classification accuracy results indicate comparable performances of the two methods. Image reconstruction and log-likelihood results show improved performance of the CD-based method. It is shown that the samples obtained from quantum annealer can be used to train an RBM on a 64-bit “bars and stripes” dataset with classification performance similar to an RBM trained with CD. Though training based on CD showed improved learning performance, training using a quantum annealer could be useful as it eliminates computationally expensive MCMC steps of CD.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
NASA NTRS2020

Image Moment-Based Extended Object Tracking for Complex Motions

A novel image moment-based model for shape estimation and tracking of an extended target moving with a complex trajectory is presented. The proposed extended object tracking algorithm is based on multiple noisy measurement points sampled from the target at each time step. The shape of the object, approximated by an ellipse, is estimated using a combination of image moments. Dynamic models of image moments for constant velocity and coordinated turn motions are mathematically derived. An unscented Kalman filter - interacting multiple model (UKF-IMM) method is used to track the object and estimate its shape. A likelihood function based on average log-likelihood is derived for the IMM filter. Simulation results of the proposed UKF-IMM algorithm with the image momentbased models are presented that show the estimation of the shape of the object moving in a complex trajectory. The intersection over union (IoU), and the root mean square errors (RMSEs) of the position and velocity of the centroid of the ellipse are used as metrics. The comparison results of the proposed algorithm with a benchmark algorithm from literature based on the IoU and RMSE metrics are presented.

Extended object tracking↗
NASA NTRS2022

Deep Interacting Multiple Model Filtering

In this paper, a deep learning-based multiple model estimation framework is presented for the state estimation of hybrid dynamical systems from high dimensional observations such as camera images. A low dimensional vector which represents the measurement of the latent dynamical system and its corresponding variance are learned using a deep encoder neural network. An Interacting Multiple Model (IMM) filter is used to generate the latent state estimates and covariances using multiple dynamical models, which can be learned using backpropagation through time. The state estimates of the dynamical system and the corresponding covariance matrix are generated from the latent state estimates and covariance using a deep decoder neural network. The whole network is trained in an end-to-end manner using a loss function which minimizes the negative log-likelihood of the neural network parameters. Simulation results are presented using a 2D bouncing ball example and estimation error statistics are computed which demonstrates the accuracy and consistency of the estimation.

Ghananeel Rotithor↗
NASA NTRSDate not supplied

Structured Covariance Gaussian Networks for Orion Crew Module Aerodynamic Uncertainty Quantification

In this paper we propose a new approach for nonlinear regression and uncertainty quantification. The method is based on a pair of neural networks which parameterize mean and dense covariance functions of a multivariate Gaussian process, trained together to maximize the log-likelihood of observing the given data. The covariance matrix is made positive definite at every input by construction. We also propose a sampling approach that produces viable surrogate function realizations from the Gaussian process. We call the proposed model a Structured Covariance Gaussian Network (SCGN). We illustrate the use of SCGNs for learning an aerodynamic response surface with built-in uncertainty for the Orion crew module. We find that SCGN provides an efficient and systematic way to learn nonlinear functional relationships and dense covariances. We compare results to a baseline Gaussian process regressor and observe that the SCGN provides comparable uncertainty descriptions with improved scalability to dataset size. The sample functions generated by SCGN are fast to evaluate online and are therefore convenient for use in trajectory simulations. These results suggest that SCGN may be a viable computational method for aerodynamic uncertainty quantification.

machine learning↗
NASA NTRSDate not supplied

Structured Covariance Gaussian Networks for Orion Crew Module Aerodynamic Uncertainty Quantification

In this paper we propose a new approach for nonlinear regression and uncertainty quantification. The method is based on a pair of neural networks which parameterize mean and dense covariance functions of a multivariate Gaussian process, trained together to maximize the log-likelihood of observing the given data. The covariance matrix is made positive definite at every input by construction. We also propose a sampling approach that produces viable surrogate function realizations from the Gaussian process. We call the proposed model a Structured Covariance Gaussian Network (SCGN). We illustrate the use of SCGNs for learning an aerodynamic response surface with built-in uncertainty for the Orion crew module. We find that SCGN provides an efficient and systematic way to learn nonlinear functional relationships and dense covariances. We compare results to a baseline Gaussian process regressor and observe that the SCGN provides comparable uncertainty descriptions with improved scalability to dataset size. The sample functions generated by SCGN are fast to evaluate online and are therefore convenient for use in trajectory simulations. These results suggest that SCGN may be a viable computational method for aerodynamic uncertainty quantification.

machine learning↗
NASA NTRS2018

Random Variables with Moment-Matching Staircase Density Functions

This paper proposes a family of random variables for uncertainty modeling. The variables of interest have a bounded support set, and prescribed values for the first four moments. We present the feasibility conditions for the existence of any of such variables, and propose a class of variables that conforms to such constraints. This class is called staircase because the density of its members is a piecewise constant function. Convex optimization is used to calculate their distributions according to several optimality criteria, including maximal entropy and maximal log-likelihood. The flexibility and efficiency of staircases enable modeling phenomena having a possibly skewed and/or multimodal response at a low computational cost. Furthermore, we provide a means to account for the uncertainty in the distribution caused by estimating staircases from data. These ideas are illustrated by generating empirical staircase predictor models. We consider the case in which the predictor matches the sample moments exactly (a setting applicable to large datasets), as well as the case in which the predictor accounts for the sampling error in such moments (a setting applicable to sparse datasets). A predictor model for the dynamics of an aeroelastic airfoil subject to flutter instability is used as an example. The resulting predictor not only describes the system's response accurately, but also enables carrying out a risk analysis for safe flight.

Luis G. Crespo↗
DOE OSTI2021

Reconstructing porous media using generative flow networks

One area of intense scientific interest for the study of sandstones, carbonates, and shale at the pore scale is the use of limited image and petrophysical data to generate multiple realizations of a rock’s pore structure. Such images aid efforts to quantify uncertainty in petrophysical properties, including porosity–permeability transforms. We develop and evaluate a deep learning-based method to synthesize porous media volumes using a so-called generative flow model trained on x-ray microscope images of rock texture and pore structure. These models are optimized on a log-likelihood objective and they synthesize large and realistic three-dimensional images. Further, we demonstrate the rapid generation of sandstone image volumes that display realism as gauged by quantitative comparison of topological features using Minkowski functionals of porosity, specific surface area, and the Euler–Poincaré characteristic (i.e., pore connectivity). We also evaluate the single-phase permeability using Navier–Stokes and lattice Boltzmann methods and show that transport properties of the generated samples match measured trends.

58 GEOSCIENCES↗
DOE OSTI2022

Search For Excited Cascade Hypersons (¿*-) Using the CLAS12 Spectrometer at Jefferson Laboratory

The number of experimentally observed doubly strange Cascade states to date is far fewer than has been predicted theoretically. The CLAS12 Very Strange physics program (E12- 11-005A) at the Thomas Jefferson National Accelerator Facility aims to study the electroproduction of these states using the newly upgraded CEBAF Large Acceptance Spectrometer for 12 GeV (CLAS12) in experimental Hall B. In this project, the reaction ep ! e0K+K+??? ! e0K+K+K?(?=?0) is studied using CLAS12 Run Group A (RG-A) data sets taken by impinging electron beams of 10:2 and 10:6 GeV energies on an LH2 target. Scattered electrons are detected with either the Forward Detector (FD), covering a polar angle range of 5? to 35? to study electroproduction, or with the Forward Tagger (FT), covering a polar angle range of 2:5? to 4:5?, to study quasi-real photoproduction. The CLAS12 detector with nearly a 4? solid angle coverage is used to detect scattered electrons and charged kaons in the ?final state. ?=?0 hyperons are reconstructed using the missing mass technique to explore intermediate doubly-strange hyperons (???) that decay to K? and ?=?0. No statistically significant ??? states other than the ???(1530) were found in the missing mass spectra based on the currently available statistics in the CLAS12-FD acceptance only. A maximum log-likelihood method is implemented to determine the statistical significance and the upper limits on the ???(1820) electroproduction cross section by constructing 95% confidence level boundaries on the ???(1820) yields. Additional research was conducted to investigate the upper limit electroproduction cross section of the reaction ep ! e0K+K+??? ! e0K+K+K?(?=?0) as a function of the electroproduced ??? mass and the differential cross section ( d?dMM(e0K+K+) ) as a function of missing mass MM(e0K+K+) for electroproduction and quasi-real photoproduction processes.

Khanal, Achyut↗
NASA NTRS2009

Hardware Implementation of Serially Concatenated PPM Decoder

A prototype decoder for a serially concatenated pulse position modulation (SCPPM) code has been implemented in a field-programmable gate array (FPGA). At the time of this reporting, this is the first known hardware SCPPM decoder. The SCPPM coding scheme, conceived for free-space optical communications with both deep-space and terrestrial applications in mind, is an improvement of several dB over the conventional Reed-Solomon PPM scheme. The design of the FPGA SCPPM decoder is based on a turbo decoding algorithm that requires relatively low computational complexity while delivering error-rate performance within approximately 1 dB of channel capacity. The SCPPM encoder consists of an outer convolutional encoder, an interleaver, an accumulator, and an inner modulation encoder (more precisely, a mapping of bits to PPM symbols). Each code is describable by a trellis (a finite directed graph). The SCPPM decoder consists of an inner soft-in-soft-out (SISO) module, a de-interleaver, an outer SISO module, and an interleaver connected in a loop (see figure). Each SISO module applies the Bahl-Cocke-Jelinek-Raviv (BCJR) algorithm to compute a-posteriori bit log-likelihood ratios (LLRs) from apriori LLRs by traversing the code trellis in forward and backward directions. The SISO modules iteratively refine the LLRs by passing the estimates between one another much like the working of a turbine engine. Extrinsic information (the difference between the a-posteriori and a-priori LLRs) is exchanged rather than the a-posteriori LLRs to minimize undesired feedback. All computations are performed in the logarithmic domain, wherein multiplications are translated into additions, thereby reducing complexity and sensitivity to fixed-point implementation roundoff errors. To lower the required memory for storing channel likelihood data and the amounts of data transfer between the decoder and the receiver, one can discard the majority of channel likelihoods, using only the remainder in operation of the decoder. This is accomplished in the receiver by transmitting only a subset consisting of the likelihoods that correspond to time slots containing the largest numbers of observed photons during each PPM symbol period. The assumed number of observed photons in the remaining time slots is set to the mean of a noise slot. In low background noise, the selection of a small subset in this manner results in only negligible loss. Other features of the decoder design to reduce complexity and increase speed include (1) quantization of metrics in an efficient procedure chosen to incur no more than a small performance loss and (2) the use of the max-star function that allows sum of exponentials to be computed by simple operations that involve only an addition, a subtraction, and a table lookup. Another prominent feature of the design is a provision for access to interleaver and de-interleaver memory in a single clock cycle, eliminating the multiple clock-cycle latency characteristic of prior interleaver and de-interleaver designs.

Moision, Bruce↗