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 325 records · Page 18
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

Seeing through noise in power laws

Despite widespread claims of power laws across the natural and social sciences, evidence in data is often equivocal. Modern data and statistical methods reject even classic power laws such as Pareto’s law of wealth and the Gutenberg–Richter law for earthquake magnitudes. We show that the maximum-likelihood estimators and Kolmogorov–Smirnov (K-S) statistics in widespread use are unexpectedly sensitive to ubiquitous errors in data such as measurement noise, quantization noise, heaping and censorship of small values. This sensitivity causes spurious rejection of power laws and biases parameter estimates even in arbitrarily large samples, which explains inconsistencies between theory and data. We show that logarithmic binning by powers of λ > 1 attenuates these errors in a manner analogous to noise averaging in normal statistics and that λ thereby tunes a trade-off between accuracy and precision in estimation. Binning also removes potentially misleading within-scale information while preserving information about the shape of a distribution over powers of λ, and we show that some amount of binning can improve sensitivity and specificity of K-S tests without any cost, while more extreme binning tunes a trade-off between sensitivity and specificity. We therefore advocate logarithmic binning as a simple essential step in power-law inference.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Tensor Decompositions for Count Data that Leverage Stochastic and Deterministic Optimization

There is growing interest to extend low-rank matrix decompositions to multi-way arrays, or tensors. One fundamental low-rank tensor decomposition is the canonical polyadic decomposition (CPD). The challenge of fitting a low-rank, nonnegative CPD model to Poisson-distributed count data is of particular interest. Several popular algorithms use local search methods to approximate the global maximum likelihood estimator from local minima. Simultaneously, a recent trend in theoretical computer science and numerical linear algebra leverages randomization to solve very large, hard problems. The typical approach is to use randomization for a fast approximation and determinism for refinement to yield effective algorithms with theoretical guarantees. Two popular algorithms for Poisson CPD reflect that emergent dichotomy: CP Alternating Poisson Regression is a deterministic algorithm and Generalized Canonical Polyadic decomposition makes use of stochastic algorithms in several variants. This work extends recent work to develop two new methods that leverage randomized and deterministic algorithms for improved accuracy and performance.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

Poisson-response Tensor-on-Tensor Regression and Applications

We introduce Poisson-response tensor-on-tensor regression (PToTR), a novel regression framework designed to handle tensor responses composed element-wise of random Poisson-distributed counts. Tensors, or multi-dimensional arrays, composed of counts are common data in fields such as inter national relations, social networks, epidemiology, and medical imaging, where events occur across multiple dimensions like time, location, and dyads. PToTR accommodates such tensor responses alongside tensor covariates, providing a versatile tool for multi dimensional data analysis. We propose algorithms for maximum likelihood estimation under a canonical polyadic (CP) structure on the regression coefficient tensor that satisfy the positivity of Poisson parameters and then provide an initial theoretical error analysis for PToTR estimators. We also demonstrate the utility of PToTR through three concrete applications: longitudinal data analysis of the Integrated Crisis Early Warning System database, positron emission tomography (PET) image reconstruction, and change-point detection of communication patterns in longitudinal dyadic data. These applications highlight the versatility of PToTR in addressing complex, structured count data across various domains.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2017

The Multi-Mission Maximum Likelihood framework (3ML)

The age of multi-wavelength and multi-messenger astronomy has arrived and with it, new tools are needed to analyze data from multiple instruments properly and with ease. The Multi-Mission Maximum Likelihood framework (3ML) provides this functionality via the novel use of instrument plugins which allow for every instrument’s unique data to be treated independently with an appropriate likelihood. Under the 3ML framework, users can design plugins that handle instrument specific data routines transparently in the background. When multiple instruments are used together, their independent likelihoods are treated under a common minimization or Bayesian sampling framework. 3ML provides a multitude of minimization algorithm for maximum likelihood estimation (MLE) as well as several popular Bayesian posterior samplers. The entire framework is provided via a modern Python interface providing the user with a modern and easily transportable analysis framework well suited for modern astronomy. New models can be added easily. It is also possible to perform time-energy modeling. We present the framework and its main functionalities.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2021

Multiscale Models for Fibril Formation: Rare Events Methods, Microkinetic Models, and Population Balances

Amyloid fibrils are thought to grow by a two-step dock-lock mechanism. However, previous simulations of fibril formation (i) overlook the bi-molecular nature of the docking step and obtain rates with first-order units, or (ii) superimpose the docked and locked states when computing the potential of mean force for association and thereby muddle the docking and locking steps. Here, we developed a simple microkinetic model with separate locking and docking steps and with the appropriate concentration dependences for each step. We constructed a simple model comprised of chiral dumbbells that retains qualitative aspects of fibril formation. We used rare events methods to predict separate docking and locking rate constants for the model. The rate constants were embedded in the microkinetic model, with the microkinetic model embedded in a population balance model for “bottom-up” multiscale fibril growth rate predictions. These were compared to “top-down” results using simulation data with the same model and multiscale framework to obtain maximum likelihood estimates of the separate lock and dock rate constants. We used the same procedures to extract separate docking and locking rate constants from experimental fibril growth data. Our multiscale strategy, embedding rate theories, and kinetic models in conservation laws should help to extract docking and locking rate constants from experimental data or long molecular simulations with correct units and without compromising the molecular description.

Shayesteh Zadeh, Armin (ORCID:0000000299896710)↗
DOE OSTI2021

Cardinality-Regularized Hawkes-Granger Model

We propose a new sparse Granger-causal learning framework for temporal event data. We focus on a specific class of point processes called the Hawkes process. We begin by pointing out that most of the existing sparse causal learning algorithms for the Hawkes process suffer from a singularity in maximum likelihood estimation. As a result, their sparse solutions can appear only as numerical artifacts. In this paper, we propose a mathematically well-defined sparse causal learning framework based on a cardinality-regularized Hawkes process, which remedies the pathological issues of existing approaches. We leverage the proposed algorithm for the task of instance-wise causal event analysis, where sparsity plays a critical role. We validate the proposed framework with two real use-cases, one from the power grid and the other from the cloud data center management domain.

Ide, Tsuyoshi↗
NASA NTRS1967

Computer program reduces calculation time of normal response functions

FORTRAN 2 computer program rapidly calculates parameters of maximum likelihood estimates from sensitivity experiment data populations. The program uses the Newton-Raphson iterative procedure to calculate the mean and standard deviation of portions of the cumulative normal response function.

Alexander, M. J.↗
NASA NTRS1971

The application of signal detection theory to optics

The restoration of images focused on a photosensitive surface is treated from the standpoint of maximum likelihood estimation, taking into account the Poisson distributions of the observed data, which are the numbers of photoelectrons from various elements of the surface. A detector of an image focused on such a surface utilizes a certain linear combination of those numbers as the optimum detection statistic. Methods for calculating the false alarm and detection probabilities are proposed. It is shown that measuring noncommuting observables in an ideal quantum receiver cannot yield a lower Bayes cost than that attainable by a system measuring only commuting observables.

Helstrom, C. W.↗
NASA NTRS1971

A system debugging model.

Consideration of the nature of the 'debugging' process applied to a new complex system during the initial period of its life. During this period failures and errors are corrected as they occur, with resulting improvement in the subsequent performance of the system. One mathematical idealization of this process leads to the assumption that system failure rate is decreasing with time. In practice, the debugging phase is considered completed when the failure rate reaches an equilibrium or constant value. Models are formulated for this phenomenon. Maximum likelihood estimates are obtained for relevant failure rate functions and for the end of the debugging period. A conservative upper confidence bound on the stable failure rate is obtained.

Barlow, R. E.↗
NASA NTRS1972

Two stochastic models useful in petroleum exploration

A model of the petroleum exploration process that tests empirically the hypothesis that at an early stage in the exploration of a basin, the process behaves like sampling without replacement is proposed along with a model of the spatial distribution of petroleum reserviors that conforms to observed facts. In developing the model of discovery, the following topics are discussed: probabilitistic proportionality, likelihood function, and maximum likelihood estimation. In addition, the spatial model is described, which is defined as a stochastic process generating values of a sequence or random variables in a way that simulates the frequency distribution of areal extent, the geographic location, and shape of oil deposits

Kaufman, G. M.↗
NASA NTRS1974

An analysis and demonstration of clock synchronization by VLBI

A prototype of a semi-real time system for synchronizing the Deep Space Net station clocks by radio interferometry was successfully demonstrated on August 30, 1972. The system utilized an approximate maximum likelihood estimation procedure for processing the data, thereby achieving essentially optimum time sync estimates for a given amount of data, or equivalently, minimizing the amount of data required for reliable estimation. Synchronization accuracies as good as 100 ns rms were achieved between Deep Space Stations 11 and 12, both at Goldstone, Calif. The accuracy can be improved by increasing the system bandwidth until the fundamental limitations due to baseline and source position uncertainties and atmospheric effects are reached. These limitations are under 10 ns for transcontinental baselines.

Hurd, W. J.↗
NASA NTRS1972

An analysis and demonstration of clock synchronization by VLBI

A prototype of a semireal-time system for synchronizing the DSN station clocks by radio interferometry was successfully demonstrated. The system utilized an approximate maximum likelihood estimation procedure for processing the data, thereby achieving essentially optimum time synchronization estimates for a given amount of data, or equivalently, minimizing the amount of data required for reliable estimation. Synchronization accuracies as good as 100 nsec rms were achieved between DSS 11 and DSS 12, both at Goldstone, California. The accuracy can be improved by increasing the system bandwidth until the fundamental limitations due to position uncertainties of baseline and source and atmospheric effects are reached. These limitations are under ten nsec for transcontinental baselines.

Hurd, W. J.↗
NASA NTRS1975

Flight-Determined Stability and Control Derivatives for an Executive Jet Transport

A modified maximum likelihood estimation (MMLE) technique which included a provision for including a priori information about unknown parameters was used to determine the aerodynamic derivatives of the Lockheed JetStar airplane. Two hundred sixty-five maneuvers were performed with the JetStar airplane, which was modified to include direct lift controls, to obtain lateral-directional and longitudinal derivatives. Data were obtained at altitudes of 3048 meters, 6096 meters, and 9144 meters (10,000 feet, 20,000 feet, and 30,000 feet) and over an angle of attack range from approximately 3 deg to 13 deg and a Mach number range from 0.25 to 0.75. Side force generators were installed and tested in 87 maneuvers to determine their effectiveness and their effect on the other derivatives. Lateral-directional data for four flight conditions were analyzed without using a priori information to assess the effect of this feature on the results. The MMLE method generally gave consistent (repeatable) estimates of the derivatives, with the exception of the rolling moment due to yaw rate, which showed large variances.

Smith, H. J.↗
NASA NTRS1975

Stability and control derivatives of the T-37B airplane

Subsonic stability and control derivatives were determined by a modified maximum likelihood estimator from flight data for the longitudinal and lateral-directional modes of the T-37B airplane. Data from two flights, in which 166 stability and control maneuvers were performed, were used in the determination. The configurations investigated were: zero flaps, gear up; half flaps, gear up; full flaps, gear up; and zero flaps, gear down.

Shafer, M. F.↗
NASA NTRS1975

UHMLE: Program description user guide

A program which computes maximum likelihood estimates for the general normal mixture is introduced. The program allows the user to fix any subsets of the mixture parameters; this fixed set may be redefined at various times in the iteration process. There is considerable time saved in computing the likelihoods if the diagonal form is specified. Hence, in the early iterations the diagonal assumption might be used, changing over to the full covariance mode later in the iteration process for a more refined solution. This flexibility allows the user to choose the sequence of parameter configurations in the iteration process which he feels will optimize the computation time required as well as possibly avoid convergence to suboptimal local maxima of the likelihood function.

Cobely, W. A.↗