Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Likelihood”

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 91 records · Page 5

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↗

Estimating the Likelihood of GHG Concentration Scenarios From Probabilistic Integrated Assessment Model Simulations

The climate scenarios that form the basis for current climate risk assessments have no assigned probabilities, and this impedes the analysis of future climate risks. This paper proposes an approach to estimate the probability of carbon dioxide (CO 2 ) concentration scenarios used in key climate change modeling experiments. It computes the CO 2 emissions compatible with the concentrations prescribed by Coupled Model Intercomparison Project Phase 5 (CMIP5) and CMIP6 experiments. The distribution of these compatible cumulative emissions is interpreted as the likelihood of future emissions given a concentration pathway. Using Bayesian analysis, the probability of each pathway can be estimated from a probabilistic sample of future emissions. The approach is demonstrated with five probabilistic CO 2 emission simulation ensembles from four Integrated Assessment Models (IAM), leading to independent estimates of the likelihood of the CO 2 concentration of Representative Concentration Pathways (RCP) and Shared Socioeconomic Pathways (SSP). Results suggest that SSP5-8.5 is unlikely for the second half of the 21st century, but offer no clear consensus on which of the remaining scenarios is most likely. Estimates of likelihoods of CO 2 concentrations associated with RCP and SSP scenarios are affected by sampling errors, differences in emission sources simulated by the IAMs, and a lack of a common experimental framework for IAM simulations. These shortcomings, along with a small IAM ensemble size, limit the applicability of the results presented here. Novel joint IAM and the Earth System Model experiments are needed to deliver actionable probabilistic climate risk assessments.

54 ENVIRONMENTAL SCIENCES↗

Planck intermediate results: XVI. Profile likelihoods for cosmological parameters

We explore the 2013 Planck likelihood function with a high-precision multi-dimensional minimizer (Minuit). This allows a refinement of the CDM best-fit solution with respect to previously-released results, and the construction of frequentist confidence intervals using profile likelihoods. The agreement with the cosmological results from the Bayesian framework is excellent, demonstrating the robustness of the Planck results to the statistical methodology. We investigate the inclusion of neutrino masses, where more significant differences may appear due to the non-Gaussian nature of the posterior mass distribution. By applying the Feldman-Cousins prescription, we again obtain results very similar to those of the Bayesian methodology. However, the profile-likelihood analysis of the cosmic microwave background (CMB) combination ( Planck +WP+highL) reveals a minimum well within the unphysical negative-mass region. We show that inclusion of the Planck CMB-lensing information regularizes this issue, and provide a robust frequentist upper limit Σ$m_v$ ≤0:26 eV (95% confidence) from the CMB+lensing+BAO data combination.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

LIKELIHOOD-FREE COSMOLOGICAL INFERENCE WITH TYPE Ia SUPERNOVAE: APPROXIMATE BAYESIAN COMPUTATION FOR A COMPLETE TREATMENT OF UNCERTAINTY

Cosmological inference becomes increasingly difficult when complex data-generating processes cannot be modeled by simple probability distributions. With the ever-increasing size of data sets in cosmology, there is an increasing burden placed on adequate modeling; systematic errors in the model will dominate where previously these were swamped by statistical errors. For example, Gaussian distributions are an insufficient representation for errors in quantities like photometric redshifts. Likewise, it can be difficult to quantify analytically the distribution of errors that are introduced in complex fitting codes. Without a simple form for these distributions, it becomes difficult to accurately construct a likelihood function for the data as a function of parameters of interest. Approximate Bayesian computation (ABC) provides a means of probing the posterior distribution when direct calculation of a sufficiently accurate likelihood is intractable. ABC allows one to bypass direct calculation of the likelihood but instead relies upon the ability to simulate the forward process that generated the data. These simulations can naturally incorporate priors placed on nuisance parameters, and hence these can be marginalized in a natural way. We present and discuss ABC methods in the context of supernova cosmology using data from the SDSS-II Supernova Survey. Assuming a flat cosmology and constant dark energy equation of state, we demonstrate that ABC can recover an accurate posterior distribution. Finally, we show that ABC can still produce an accurate posterior distribution when we contaminate the sample with Type IIP supernovae.

79 ASTRONOMY AND ASTROPHYSICS↗

Caching and interpolated likelihoods: accelerating cosmological Monte Carlo Markov chains

We describe a novel approach to accelerating Monte Carlo Markov Chains. Our focus is cosmological parameter estimation, but the algorithm is applicable to any problem for which the likelihood surface is a smooth function of the free parameters and computationally expensive to evaluate. We generate a high-order interpolating polynomial for the log-likelihood using the first points gathered by the Markov chains as a training set. This polynomial then accurately computes the majority of the likelihoods needed in the latter parts of the chains. We implement a simple version of this algorithm as a patch (InterpMC) to CosmoMC and show that it accelerates parameter estimatation by a factor of between two and four for well-converged chains. The current code is primarily intended as a ''proof of concept'', and we argue that there is considerable room for further performance gains. Unlike other approaches to accelerating parameter fits, we make no use of precomputed training sets or special choices of variables, and InterpMC is almost entirely transparent to the user.

79 ASTRONOMY AND ASTROPHYSICS↗

Improved renormalization group computation of likelihood functions for cosmological data sets

Evaluation of likelihood functions for cosmological large scale structure data sets (including CMB, galaxy redshift surveys, etc.) naturally involves marginalization, i.e., integration, over an unknown underlying random signal field. Recently, I showed how a renormalization group method can be used to carry out this integration efficiently by first integrating out the smallest scale structure, i.e., localized structure on the scale of differences between nearby data cells, then combining adjacent cells in a coarse graining step, then repeating this process over and over until all scales have been integrated. Here I extend the formulation in several ways in order to reduce the prefactor on the method's linear scaling with data set size. The key improvement is showing how to integrate out the difference between specific adjacent cells before summing them in the coarse graining step, compared to the original formulation in which small-scale fluctuations were integrated more generally. I suggest some other improvements in details of the scheme, including showing how to perform the integration around a maximum likelihood estimate for the underlying random field. In the end, an accurate likelihood computation for a million-cell Gaussian test data set runs in two minutes on my laptop, with room for further optimization and straightforward parallelization.

79 ASTRONOMY AND ASTROPHYSICS↗

Planck integrated Sachs-Wolfe-lensing likelihood and the CMB temperature

Here we present a new Planck CMB lensing-CMB temperature cross-correlation likelihood that can be used to constrain cosmology via the integrated Sachs-Wolfe (ISW) effect. CMB lensing is an excellent tracer of ISW, and we use the latest PR4 Planck data maps and lensing reconstruction to produce the first public Planck likelihood to constrain this signal. We demonstrate the likelihood by constraining the CMB background temperature from Planck data alone, where the ISW-lensing cross-correlation is a powerful way to break the geometric degeneracy, substantially improving constraints from the CMB and lensing power spectra alone.

79 ASTRONOMY AND ASTROPHYSICS↗

Free kick instead of cross-validation in maximum-likelihood refinement of macromolecular crystal structures

The maximum-likelihood free-kick target, which calculates model error estimates from the work set and a randomly displaced model, proved superior in the accuracy and consistency of refinement of crystal structures compared with the maximum-likelihood cross-validation target, which calculates error estimates from the test set and the unperturbed model. The refinement of a molecular model is a computational procedure by which the atomic model is fitted to the diffraction data. The commonly used target in the refinement of macromolecular structures is the maximum-likelihood (ML) function, which relies on the assessment of model errors. The current ML functions rely on cross-validation. They utilize phase-error estimates that are calculated from a small fraction of diffraction data, called the test set, that are not used to fit the model. An approach has been developed that uses the work set to calculate the phase-error estimates in the ML refinement from simulating the model errors via the random displacement of atomic coordinates. It is called ML free-kick refinement as it uses the ML formulation of the target function and is based on the idea of freeing the model from the model bias imposed by the chemical energy restraints used in refinement. This approach for the calculation of error estimates is superior to the cross-validation approach: it reduces the phase error and increases the accuracy of molecular models, is more robust, provides clearer maps and may use a smaller portion of data for the test set for the calculation of R{sub free} or may leave it out completely.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Maximum likelihood positioning algorithm for high-resolution PET scanners

Purpose: In high-resolution positron emission tomography (PET), lightsharing elements are incorporated into typical detector stacks to read out scintillator arrays in which one scintillator element (crystal) is smaller than the size of the readout channel. In order to identify the hit crystal by means of the measured light distribution, a positioning algorithm is required. One commonly applied positioning algorithm uses the center of gravity (COG) of the measured light distribution. The COG algorithm is limited in spatial resolution by noise and intercrystal Compton scatter. The purpose of this work is to develop a positioning algorithm which overcomes this limitation. Methods: The authors present a maximum likelihood (ML) algorithm which compares a set of expected light distributions given by probability density functions (PDFs) with the measured light distribution. Instead of modeling the PDFs by using an analytical model, the PDFs of the proposed ML algorithm are generated assuming a single-gamma-interaction model from measured data. The algorithm was evaluated with a hot-rod phantom measurement acquired with the preclinical HYPERION II {sup D} PET scanner. In order to assess the performance with respect to sensitivity, energy resolution, and image quality, the ML algorithm was compared to a COG algorithm which calculates the COG from a restricted set of channels. The authors studied the energy resolution of the ML and the COG algorithm regarding incomplete light distributions (missing channel information caused by detector dead time). Furthermore, the authors investigated the effects of using a filter based on the likelihood values on sensitivity, energy resolution, and image quality. Results: A sensitivity gain of up to 19% was demonstrated in comparison to the COG algorithm for the selected operation parameters. Energy resolution and image quality were on a similar level for both algorithms. Additionally, the authors demonstrated that the performance of the ML algorithm is less prone to missing channel information. A likelihood filter visually improved the image quality, i.e., the peak-to-valley increased up to a factor of 3 for 2-mm-diameter phantom rods by rejecting 87% of the coincidences. A relative improvement of the energy resolution of up to 12.8% was also measured rejecting 91% of the coincidences. Conclusions: The developed ML algorithm increases the sensitivity by correctly handling missing channel information without influencing energy resolution or image quality. Furthermore, the authors showed that energy resolution and image quality can be improved substantially by rejecting events that do not comply well with the single-gamma-interaction model, such as Compton-scattered events.

60 APPLIED LIFE SCIENCES↗

Likelihood Functions for Supersymmetric Observables in Frequentist Analyses of the CMSSM and NUHM1

On the basis of frequentist analyses of experimental constraints from electroweak precision data, g-2, B physics and cosmological data, we investigate the parameters of the constrained MSSM (CMSSM) with universal soft supersymmetry-breaking mass parameters, and a model with common non-universal Higgs masses (NUHM1). We present chi^2 likelihood functions for the masses of supersymmetric particles and Higgs bosons, as well as b to s gamma, b to mu mu and the spin-independent dark matter scattering cross section. In the CMSSM we find preferences for sparticle masses that are relatively light. In the NUHM1 the best-fit values for many sparticle masses are even slightly smaller, but with greater uncertainties. The likelihood functions for most sparticle masses are cut off sharply at small masses, in particular by the LEP Higgs mass constraint. Both in the CMSSM and the NUHM1, the coannihilation region is favoured over the focus-point region at about the 3-sigma level, largely but not exclusively because of g-2. Many sparticle masses are highly correlated in both the CMSSM and NUHM1, and most of the regions preferred at the 95% C.L. are accessible to early LHC running. Some slepton and chargino/neutralino masses should be in reach at the ILC. The masses of the heavier Higgs bosons should be accessible at the LHC and the ILC in portions of the preferred regions in the (M_A, tan beta) plane. In the CMSSM, the likelihood function for b to mu mu is peaked close to the Standard Model value, but much larger values are possible in the NUHM1. We find that values of the DM cross section > 10^{-10} pb are preferred in both the CMSSM and the NUHM1. We study the effects of dropping the g-2, b to s gamma, relic density and M_h constraints.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Point Source Likelihood Framework (psl) v1.0

PSL is a software framework for localizing radiation point sources in a 3D environment using a freely moving radiation detector system. The framework assumes a single point source of radiation is in the environment and uses known detector positions and orientations along with their radiation measurements to find the most likely location of the point source. This approach to the problem enables analyses that improve upon the capability of a standard Maximum Likelihood Expectation Maximization (MLEM) approach, such as setting a source detection threshold, finding a likelihood interval for the position of the source, finding a likelihood interval for the activity (intensity) of the source, determining whether more than one point source may be present, and generating a map of Minimum Detectable Activity (MDA) for a given source type. Applications include radiological and nuclear security, consequence management, decontamination and decommissioning of nuclear facilities, and radiological emergency response.

Joshi, Tenzing↗

New machine learning techniques for simulation-based inference: InferoStatic nets, kernel score estimation, and kernel likelihood ratio estimation

We propose an intuitive, machine-learning approach to multiparameter inference, dubbed the InferoStatic Networks (ISN) method, to model the score and likelihood ratio estimators in cases when the probability density can be sampled but not computed directly. The ISN uses a backend neural network that models a scalar function called the inferostatic potential \varphi φ . In addition, we introduce new strategies, respectively called Kernel Score Estimation (KSE) and Kernel Likelihood Ratio Estimation (KLRE), to learn the score and the likelihood ratio functions from simulated data. We illustrate the new techniques with some toy examples and compare to existing approaches in the literature. We mention en passant some new loss functions that optimally incorporate latent information from simulations into the training procedure.

Kong, Kyoungchul↗

Procedure for estimating stability and control parameters from flight test data by using maximum likelihood methods employing a real-time digital system

A maximum likelihood parameter estimation procedure and program were developed for the extraction of the stability and control derivatives of aircraft from flight test data. Nonlinear six-degree-of-freedom equations describing aircraft dynamics were used to derive sensitivity equations for quasilinearization. The maximum likelihood function with quasilinearization was used to derive the parameter change equations, the covariance matrices for the parameters and measurement noise, and the performance index function. The maximum likelihood estimator was mechanized into an iterative estimation procedure utilizing a real time digital computer and graphic display system. This program was developed for 8 measured state variables and 40 parameters. Test cases were conducted with simulated data for validation of the estimation procedure and program. The program was applied to a V/STOL tilt wing aircraft, a military fighter airplane, and a light single engine airplane. The particular nonlinear equations of motion, derivation of the sensitivity equations, addition of accelerations into the algorithm, operational features of the real time digital system, and test cases are described.

Grove, R. D.↗

Convolutional codes. II - Maximum-likelihood decoding. III - Sequential decoding

Maximum-likelihood decoding is characterized as the determination of the shortest path through a topological structure called a trellis. Aspects of code structure are discussed along with questions regarding maximum-likelihood decoding on memoryless channels. A general bounding technique is introduced. The technique is used to obtain asymptotic bounds on the probability of error for maximum-likelihood decoding and list-of-2 decoding. The basic features of sequential algorithms are discussed along with a stack algorithm, questions of computational distribution, and the martingale approach to computational bounds.

Forney, G. D., Jr.↗

Maximum likelihood estimates of lift and drag characteristics obtained from dynamic aircraft maneuvers

A maximum likelihood estimation method for obtaining lift and drag characteristics from dynamic flight maneuvers was investigated. This paper describes the method and compares the estimates of lift and drag obtained by using the method with estimates obtained from wind-tunnel tests and from established methods for obtaining estimates from flight data. In general, the lift and drag coefficients extracted from dynamic flight maneuvers by the maximum likelihood estimation technique are in good agreement with the estimates obtained from the wind-tunnel tests and the other methods. When maneuvers that met the requirements of both flight methods were analyzed, the results of each method were nearly the same. The maximum likelihood estimation technique showed promise in terms of estimating lift and drag characteristics from dynamic flight maneuvers. Further studies should be made to assess the best mathematical model and the most desirable type of dynamic maneuver to get the highest quality results from this technique.

Iliff, K. W.↗

The numerical evaluation of maximum-likelihood estimates of the parameters for a mixture of normal distributions from partially identified samples

Likelihood equations determined by the two types of samples which are necessary conditions for a maximum-likelihood estimate were considered. These equations suggest certain successive approximations iterative procedures for obtaining maximum likelihood estimates. The procedures, which are generalized steepest ascent (deflected gradient) procedures, contain those of Hosmer as a special case.

Walker, H. F.↗

Mixture densities, maximum likelihood, and the EM algorithm

The problem of estimating the parameters which determine a mixture density is reviewed as well as maximum likelihood estimation for it. A particular iterative procedure for numerically approximating maximum likelihood estimates for mixture density problems is considered. This EM algorithm, is a specialization to the mixture density context of a general algorithm of the same name used to approximate maximum likelihood estimates for incomplete data problems. The formulation and theoretical and practical properties of the EM algorithm for mixture densities are discussed focussing in particular on mixtures of densities from exponential families.

Redner, R. A.↗

Maximum likelihood estimation of turbulence spectrum parameters

Estimation of the integral scale and intensity of a generic turbulence record is treated as a statistical problem of parameter estimation. Properties of parameter estimators and the method of maximum likelihood are reviewed. Likelihood equations are derived for estimation of the integral scale and intensity applicable to a general class of turbulence spectra that includes the von Karman and Dryden transverse and longitudinal spectra as special cases. The method is extended to include the Bullen transverse and longitudinal spectra. Coefficients of variation are given for maximum likelihood estimates of the integral scale and intensity of the von Karman spectra. Application of the method is illustrated by estimating the integral scale and intensity of an atmospheric turbulence vertical velocity record assumed to be governed by the von Karman transverse spectrum.

Mark, W. D.↗