Search NASA⌕ Search

SEARCH · Search NASA

Results for “Maximum Likelihood Estimation”

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.

At least 181 records · Page 10
All sourcesNASA NTRSDOE OSTI
DOE OSTI2022

Likelihood Maximization and Moment Matching in Low SNR Gaussian Mixture Models

We derive an asymptotic expansion for the log-likelihood of Gaussian mixture models (GMMs) with equal covariance matrices in the low signal-to-noise regime. The expansion reveals an intimate connection between two types of algorithms for parameter estimation: the method of moments and likelihood optimizing algorithms such as Expectation-Maximization (EM). We show that likelihood optimization in the low SNR regime reduces to a sequence of least squares optimization problems that match the moments of the estimate to the ground truth moments one by one. This connection is a stepping stone towards the analysis of EM and maximum likelihood estimation in a wide range of models. A motivating application for the study of low SNR mixture models is cryo-electron microscopy data, which can be modeled as a GMM with algebraic constraints imposed on the mixture centers. We discuss the application of our expansion to algebraically constrained GMMs, among other example models of interest. © 2022 The Authors. Communications on Pure and Applied Mathematics published by Wiley Periodicals LLC.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1994

On determining the spectrum of primordial inhomogeneity from the COBE DMR sky maps: Results of two-year data analysis

A new technique of Fourier analysis on a cut sky has been applied to the two-year Cosmic Background Explorer (COBE) Differential Microwave Radiometer (DMR) 53 and 90 GHz sky maps. The Bayesian power spectrum estimation results are consistent with the Harrison-Zel'dovich n = 1 model. The maximum likelihood estimates of the usual parameters defining the power spectrum of primordial perturbations are n = 1.22 (1.02) and Q(sub rms-PS) = 17 (20) microK including (excluding) the quadrupole. A spectral-index-independent normalization is naturally expressed for the two-year maps in terms of the multipole amplitude a(sub 9) = 8.2 (8.3) microK (to approximately 12 sigma significance). The marginal likelihood function on n obtained by intergration with respect to a(sub 9) renders n = 1.17 +/- 0.31 (0.96 +/- 0.36).

Gorski, K. M.↗
NASA NTRS1994

Performance of mean-frequency estimators for Doppler radar and lidar

The performance of mean-frequency estimators for Doppler radar and lidar measurements of winds is presented in terms of two basic parameters: Phi, the ratio of the average signal energy per estimate to the spectral noise level; and Omega, which is proportional to the number of independent samples per estimate. For fixed Phi and Omega, the Cramer-Rao bound (CRB) (theoretical best performance) for unbiased estimators of mean frequency (normalized by the spectral width of the signal), signal power, and spectral width are essentially independent of the number of data samples M. For large Phi, the estimators of mean frequency are unbiased and the performance is independent of M. The spectral domain estimators and covariance based estimators are bounded by the approximate period of M. The spectral domain estimators and covariance based estimators are bounded by the approximate periodogram CRB. The standard deviation of the maximum-likelihood estimator approaches the exact CRB, which can be more than a factor of 2 better than the performance of the spectral domain estimators or covariance-based estimators for typical Omega. For small Phi, the estimators are biased due to the effects of the uncorrelated noise (white noise), which results in uniformly distributed 'bad' estimates. The fraction of bad estimates is a function of Phi and M with weak dependence on the parameter Omega. Simple empirical models describe the standard deviation of the good estimates and the fraction of bad estimates. For Doppler lidar and for large Phi, better performance is obtained by using many low-energy pulses instead of one pulse with the same total energy. For small Phi, the converse is true.

Frehlich, R. G.↗
NASA NTRS2022

Vision-based Velocimetry over Unknown Terrain with a Low-Noise IMU

This paper presents a novel approach to terrain-relative navigation with a visual camera and Inertial Measurement Unit (IMU). The proposed algorithm uses an Extended Kalman Filter (EKF) to combine an IMU propagated state estimate with batch correction estimates computed over a sliding window of measurements. The batch correction algorithm follows the Maximum Likelihood Estimation (MLE) approach used in other Bundle Adjustment systems. Unlike other systems, the proposed system parameterizes the state over the entire window in terms of the state at a single epoch. By ignoring IMU error over the window duration, we obtain a state epoch MLE that jointly estimates the epoch state and terrain parameters with drastically reduced computationally cost. This paper presents the general architecture which can be adapted for various state parameterizations and measurement inputs. For space applications with high-accuracy IMUs, the reduction in computational cost comes with only a modest increase in estimation errors. The increase in error is quantified via a linear covariance analysis presented in this paper. Furthermore, we present simulation results which show the applicability of this algorithm to planetary landing problems.

San Martin, A. Miguel↗
NASA NTRS1979

Parameter estimation in X-ray astronomy using maximum likelihood

Methods of estimation of parameter values and confidence regions by maximum likelihood and Fisher efficient scores starting from Poisson probabilities are developed for the nonlinear spectral functions commonly encountered in X-ray astronomy. It is argued that these methods offer significant advantages over the commonly used alternatives called minimum chi-squared because they rely on less pervasive statistical approximations and so may be expected to remain valid for data of poorer quality. Extensive numerical simulations of the maximum likelihood method are reported which verify that the best-fit parameter value and confidence region calculations are correct over a wide range of input spectra.

Wachter, K.↗
NASA NTRS1974

Radar optimization for sea surface and geodetic measurements

The efficient estimation of geoid and sea state parameters is discussed, and the optimum processing structures, including maximum likelihood estimators, and their accuracy limits are given for a model. The model accounts for random surface reflectivity, sea height, and additive noise, and allows for arbitrary radar system parameters, based on the assumption the received signal is a sample function of a normal random process. The integral equation associated with the Gaussian signal in Gaussian noise inference problem was solved. It is shown that the optimum processing is generally a mixture of coherent and incoherent integrations which may be viewed as a weighted summation of received power of the match-filtered received data. When estimates are correlated, the strongest correlation appears between geoid and asymmetry estimates, and between wave height standard deviation and reflectivity estimates.

Harger, R. O.↗
NASA NTRS1983

Cleanroom certification model

The Cleanroom software development methodology is designed to take the gamble out of product releases for both suppliers and receivers of the software. The ingredients of this procedure are a life cycle of executable product increments, representative statistical testing, and a standard estimate of the MTTF (Mean Time To Failure) of the product at the time of its release. A statistical approach to software product testing using randomly selected samples of test cases is considered. A statistical model is defined for the certification process which uses the timing data recorded during test. A reasonableness argument for this model is provided that uses previously published data on software product execution. Also included is a derivation of the certification model estimators and a comparison of the proposed least squares technique with the more commonly used maximum likelihood estimators.

Currit, P. A.↗
NASA NTRS1974

Recursive estimation of prior probabilities using the mixture approach

The problem of estimating the prior probabilities q sub k of a mixture of known density functions f sub k(X), based on a sequence of N statistically independent observations is considered. It is shown that for very mild restrictions on f sub k(X), the maximum likelihood estimate of Q is asymptotically efficient. A recursive algorithm for estimating Q is proposed, analyzed, and optimized. For the M = 2 case, it is possible for the recursive algorithm to achieve the same performance with the maximum likelihood one. For M 2, slightly inferior performance is the price for having a recursive algorithm. However, the loss is computable and tolerable.

Kazakos, D.↗
DOE OSTI2019

Dark Energy Survey Year 1 results: measurement of the galaxy angular power spectrum

We use data from the first-year observations of the DES collaboration to measure the galaxy angular power spectrum (APS), and search for its BAO feature. We test our methodology in a sample of 1800 DES Y1-like mock catalogues. We use the pseudo-C ℓ method to estimate the APS and the mock catalogues to estimate its covariance matrix. We use templates to model the measured spectra and estimate template parameters firstly from the C ℓ ’s of the mocks using two different methods, a maximum likelihood estimator and a Markov Chain Monte Carlo, finding consistent results with a good reduced χ 2 . Robustness tests are performed to estimate the impact of different choices of settings used in our analysis. Finally, we apply our method to a galaxy sample constructed from DES Y1 data specifically for LSS studies. This catalogue comprises galaxies within an effective area of 1318 deg 2 and 0.6 < z < 1.0. We find that the DES Y1 data favour a model with BAO at the 2.6σ C.L. However, the goodness of fit is somewhat poor, with χ 2 /(d.o.f.) = 1.49. We identify a possible cause showing that using a theoretical covariance matrix obtained from C ℓ ’s that are better adjusted to data results in an improved value of χ 2 /(dof) = 1.36 which is similar to the value obtained with the real-space analysis. Lastly, our results correspond to a distance measurement of D A (z eff = 0.81)/r d = 10.65 ± 0.49, consistent with the main DES BAO findings. This is a companion paper to the main DES BAO article showing the details of the harmonic space analysis.

79 ASTRONOMY AND ASTROPHYSICS↗
NASA NTRS1980

Maximum likelihood method for estimating airplane stability and control parameters from flight data in frequency domain

A frequency domain maximum likelihood method is developed for the estimation of airplane stability and control parameters from measured data. The model of an airplane is represented by a discrete-type steady state Kalman filter with time variables replaced by their Fourier series expansions. The likelihood function of innovations is formulated, and by its maximization with respect to unknown parameters the estimation algorithm is obtained. This algorithm is then simplified to the output error estimation method with the data in the form of transformed time histories, frequency response curves, or spectral and cross-spectral densities. The development is followed by a discussion on the equivalence of the cost function in the time and frequency domains, and on advantages and disadvantages of the frequency domain approach. The algorithm developed is applied in four examples to the estimation of longitudinal parameters of a general aviation airplane using computer generated and measured data in turbulent and still air. The cost functions in the time and frequency domains are shown to be equivalent; therefore, both approaches are complementary and not contradictory. Despite some computational advantages of parameter estimation in the frequency domain, this approach is limited to linear equations of motion with constant coefficients.

Klein, V.↗
DOE OSTI2019

DeepCMB: Lensing reconstruction of the cosmic microwave background with deep neural networks

Next-generation cosmic microwave background (CMB) experiments will have lower noise and therefore increased sensitivity, enabling improved constraints on fundamental physics parameters such as the sum of neutrino masses and the tensor-to-scalar ratio r. Achieving competitive constraints on these parameters requires high signal-to-noise extraction of the projected gravitational potential from the CMB maps. Standard methods for reconstructing the lensing potential employ the quadratic estimator (QE). However, the QE performs suboptimally at the low noise levels expected in upcoming experiments. Other methods, like maximum likelihood estimators (MLE), are under active development. In this work, we demonstrate reconstruction of the CMB lensing potential with deep convolutional neural networks (CNN) - ie, a ResUNet. The network is trained and tested on simulated data, and otherwise has no physical parametrization related to the physical processes of the CMB and gravitational lensing. We show that, over a wide range of angular scales, ResUNets recover the input gravitational potential with a higher signal-to-noise ratio than the QE method, reaching levels comparable to analytic approximations of MLE methods. We demonstrate that the network outputs quantifiably different lensing maps when given input CMB maps generated with different cosmologies. We also show we can use the reconstructed lensing map for cosmological parameter estimation. This application of CNN provides a few innovations at the intersection of cosmology and machine learning. First, while training and regressing on images, we predict a continuous-variable field rather than discrete classes. Second, we are able to establish uncertainty measures for the network output that are analogous to standard methods. We expect this approach to excel in capturing hard-to-model non-Gaussian astrophysical foreground and noise contributions.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2019

DeepCMB: Lensing Reconstruction of the CMB with Deep Neural Networks

Next-generation cosmic microwave background (CMB) experiments will have lower noise and therefore increased sensitivity, enabling improved constraints on fundamental physics parameters such as the sum of neutrino masses and the tensor-to-scalar ratio r. Achieving competitive constraints on these parameters requires high signal-to-noise extraction of the projected gravitational potential from the CMB maps. Standard methods for reconstructing the lensing potential employ the quadratic estimator (QE). However, the QE performs suboptimally at the low noise levels expected in upcoming experiments. Other methods, like maximum likelihood estimators (MLE), are under active development. In this work, we demonstrate reconstruction of the CMB lensing potential with deep convolutional neural networks (CNN) - ie, a ResUNet. The network is trained and tested on simulated data, and otherwise has no physical parametrization related to the physical processes of the CMB and gravitati onal lensing. We show that, over a wide range of angular scales, ResUNets recover the input gravitational potential with a higher signal-to-noise ratio than the QE method, reaching levels comparable to analytic approximations of MLE methods. We demonstrate that the network outputs quantifiably different lensing maps when given input CMB maps generated with different cosmologies. We also show we can use the reconstructed lensing map for cosmological parameter estimation. This application of CNN provides a few innovations at the intersection of cosmology and machine learning. First, while training and regressing on images, we predict a continuous-variable field rather than discrete classes. Second, we are able to establish uncertainty measures for the network output that are analogous to standard methods. We expect this approach to excel in capturing hard-to-model non-Gaussian astrophysical foreground and noise contributions.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2019

Estimation and uncertainty quantification for the output from quantum simulators

The problem of estimating certain distributions over {0, 1} d is considered here. The distribution represents a quantum system of d qubits, where there are non-trivial dependencies between the qubits. A maximum entropy approach is adopted to reconstruct the distribution from exact moments or observed empirical moments. The Robbins Monro algorithm is used to solve the intractable maximum entropy problem, by constructing an unbiased estimator of the un-normalized target with a sequential Monte Carlo sampler at each iteration. In the case of empirical moments, this coincides with a maximum likelihood estimator. A Bayesian formulation is also considered in order to quantify uncertainty a posteriori. Several approaches are proposed in order to tackle this challenging problem, according to recently developed methodologies. In particular, unbiased estimators of the gradient of the log posterior are constructed and used within a provably convergent Langevin-based Markov chain Monte Carlo method. The methods are illustrated on classically simulated output from quantum simulators.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS2022

Enabling a Larger Deep Space Mission Suite: A Deep Space Network Queuing Antenna for Demand Access

The advent of deep space small spacecraft, as exemplified by the Mars Cubesat One (MarCO), Lunar Trailblazer, Janus, the Escape and Plasma Acceleration and Dynamics Explorers (EscaPADE), and the thirteen Artemis 1 missions, opens the possibility that a much larger number of deep space spacecraft may be launched over the next 10 years and beyond. While scientifically exciting, the prospect of a (much) larger mission suite raises significant challenges for the current approach to ground stations and mission operations. We have been investigating an integrated approach for ground stations and missions operations to enable new modes of operation while maintaining the capabilities of the current operational techniques. This integrated approach is built around three core capabilities: (1) A queuing antenna that enables monitoring the status of a much larger number of spacecraft, and allows spacecraft to transmit requests for telemetry with NASA’s Deep Space Network (DSN); (2) a flexible scheduling system that expands the current DSN scheduling services to enable allocating time on DSN antennas in near real-time; and (3) a cloud-based ground data system that can be spun up and down according to how tracks are assigned by the flexible scheduling system. We shall show that an 18 meter DSN queuing antenna equipped with cyrogenic receivers would enable use of the DSN Demand Access Service for small spacecraft throughout the inner Solar System, thus providing service to a large mission suite. We first discuss the architecture of the queuing antenna and its supporting systems, including, for instance, the service required to generate the schedule for the queueing antenna (which dictates how it slews to monitor multiple spacecraft in a day of operations). Next, we describe the signaling scheme used to encode a request, which is inherited from the already operational DSN Beacon Tone Service, and describe two alternative ways to detect the incoming tone at the ground station, one based on maximum likelihood estimation (MLE), and another one based on Fast-Fourier Transfer (FFT) processing. We then use these results to estimate the maximum range at which a request can be reliably detected as a function of the spacecraft and ground station communication capabilities. Finally, the last part of this part of this paper briefly describes the prototyping effort undertaken at Morehead State University (MSU) and JPL to demonstrate the viability of this new DSN demand access. In particular, we describe the suite of tests conducted using MSU’s 21 meter ground station to validate its use a queuing antenna.

Mattle, Emily↗
NASA NTRS1977

F-8C adaptive flight control laws

Three candidate digital adaptive control laws were designed for NASA's F-8C digital flyby wire aircraft. Each design used the same control laws but adjusted the gains with a different adaptative algorithm. The three adaptive concepts were: high-gain limit cycle, Liapunov-stable model tracking, and maximum likelihood estimation. Sensors were restricted to conventional inertial instruments (rate gyros and accelerometers) without use of air-data measurements. Performance, growth potential, and computer requirements were used as criteria for selecting the most promising of these candidates for further refinement. The maximum likelihood concept was selected primarily because it offers the greatest potential for identifying several aircraft parameters and hence for improved control performance in future aircraft application. In terms of identification and gain adjustment accuracy, the MLE design is slightly superior to the other two, but this has no significant effects on the control performance achievable with the F-8C aircraft. The maximum likelihood design is recommended for flight test, and several refinements to that design are proposed.

Hartmann, G. L.↗
DOE OSTI2019

Distributions of the Mean for Lognormally Distributed Data - 19427

In environmental sciences, a distribution of the mean is often needed for right-skewed data. For example, risk assessments are often based on distributions of the mean to obtain upper confidence limits (UCLs) for analyte concentrations, and, in probabilistic performance assessment (PPA) models, parameters are given distributions rather than deterministic values, and they are usually given distributions of their mean rather than the population distributions to account for the large spatial and temporal domains to which they are applied. How this distribution of the mean is developed depends on the characteristics of the parameter's full distribution and how the parameter relates to its use. In this presentation, I will discuss the implications of a right skew in distribution development. Parameters may vary over orders of magnitude and/or may be right-skewed. The material property of saturated hydraulic conductivity, Ks, is a prime example of such a parameter. Right-skewed parameters are often assumed to follow lognormal distributions. If X follows a lognormal distribution, then log(X) follows a normal distribution. For some right-skewed parameters, practitioners go back and forth between the two forms of the data depending on their needs. For example, it may be preferred to use the untransformed version of X in a model or equation, but it is easier to visualize the parameter values on a logarithmic scale. However, distribution development for the average value of the parameter will differ depending on whether X or log(X) is modeled. This difference is due to Jensen's Inequality, a theorem which states that the expected value of a concave function of a variable is less than or equal to the same function of the expected value of the variable; a logarithm is a concave function. In other words, the mean of log(X) is always less than or equal to the log of the mean of X and therefore the mean estimated from a distribution based on log(X) is less than or equal to the mean estimated from a distribution based on X. If a parameter is naturally considered on a log scale by subject matter experts, then distributions developed for the mean of log(X) may be more consistent with the use of the parameters in the literature than distributions developed for X. The statistical decisions of distribution development will be discussed within the context of right-skewed, lognormal data. These issues include method-of-moments versus maximum likelihood estimation of the distribution parameters and the use of bootstrapping to aid in the visualization of a distribution of the mean. The details surrounding Jensen's Inequality will be explained and visualized within the context of distribution development of a right-skewed parameter such as saturated hydraulic conductivity. (authors)

54 ENVIRONMENTAL SCIENCES↗
NASA NTRS1988

Altimetry for non-Gaussian oceans - Height biases and estimation of parameters

The effects of a non-Gaussian ocean on satellite altimetry parameter estimation are discussed. The first part of this paper shows how non-Gaussian ocean parameters affect height estimation for satellites of the Seasat/Geosat/TOPEX type. In the second part, the estimation of the altimeter tracker bias and the non-Gaussian ocean parameters from the altimeter return signal is studied. A new convolution model that facilitates the deconvolution of the ocean surface specular point probability density function is introduced. Next, it is argued that it is not feasible in practice to estimate the electromagnetic bias from noisy altimeter returns. It is then shown that the least squares estimation of the surface parameters in the log-frequency domain has several advantages overestimating parameters in the time domain. The maximum likelihood estimation equations for the estimation of the waveform parameters are derived, and their solution is discussed.

Rodriguez, Ernesto↗
NASA NTRS1997

A Correlation Between Flight-Determined Derivatives and Wind-Tunnel Data for the X-24B Research Aircraft

Longitudinal and lateral-directional estimates of the aerodynamic derivatives of the X-24B research aircraft were obtained from flight data by using a modified maximum likelihood estimation method. Data were obtained over a Mach number range from 0.35 to 1.72 and over an angle of attack range from 3.5 deg. to 15.7 deg. Data are presented for a subsonic and transonic configuration. The flight derivatives were generally consistent and documented the aircraft well. The correlation between the flight data and wind-tunnel predictions is presented and discussed.

Sim, Alex G.↗