Search NASA⌕ Search

SEARCH · Search NASA

Results for “Loss 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.

At least 55 records · Page 3

A Nonlocal-Gradient Descent Method for Inverse Design in Nanophotonics

Local-gradient-based optimization approaches lack nonlocal exploration abilityrequired for escaping from local minima when searching non-convex landscapes.A directional Gaussian smoothing (DGS) approach was recently proposed in [29]and used to define a truly nonlocal gradient, referred to as the DGS gradient, inorder to enable nonlocal exploration in high-dimensional black-box optimization.Promising results show that replacing the traditional local gradient with the nonlocalDGS gradient can significantly improve the performance of gradient-based methodsin optimizing highly multi-modal loss functions. However, the current DGS methodis designed for unbounded and uncontrained optimization problems, making itinapplicable to real-world engineering optimization problems where the tuningparameters are often bounded and the loss function is usually constrained byphysical processes. In this work, we propose to extend to the DGS approachto the constrained inverse design framework in order to find better optima ofmulti-modal loss functions. A series of adaptive strategies for smoothing radiusand learning rate updating are developed to improve the computational efficiencyand robustness. Our methodology is demonstrated by an example of designing ananoscale wavelength demultiplexer, and shows superior performance compared tothe state-of-the-art approaches. By incorporating volume constraints, the optimizeddesign achieves an equivalently high performance but significantly reduces theamount of material usage.

Bi, Sirui↗

Applying Physics-Informed Neural Networks to Solve Navier–Stokes Equations for Laminar Flow around a Particle

In recent years, Physics-Informed Neural Networks (PINNs) have drawn great interest among researchers as a tool to solve computational physics problems. Unlike conventional neural networks, which are black-box models that “blindly” establish a correlation between input and output variables using a large quantity of labeled data, PINNs directly embed physical laws (primarily partial differential equations) within the loss function of neural networks. By minimizing the loss function, this approach allows the output variables to automatically satisfy physical equations without the need for labeled data. The Navier–Stokes equation is one of the most classic governing equations in thermal fluid engineering. This study constructs a PINN to solve the Navier–Stokes equations for a 2D incompressible laminar flow problem. Flows passing around a 2D circular particle are chosen as the benchmark case, and an elliptical particle is also examined to enrich the research. The velocity and pressure fields are predicted by the PINNs, and the results are compared with those derived from Computational Fluid Dynamics (CFD). Additionally, the particle drag force coefficient is calculated to quantify the discrepancy in the results of the PINNs as compared to CFD outcomes. The drag coefficient maintained an error within 10% across all test scenarios.

Hu, Beichao (ORCID:0009000163215151)↗

Train, Inform, Borrow, or Combine? Approaches to Process–Guided Deep Learning for Groundwater–Influenced Stream Temperature Prediction

Although groundwater discharge is a critical stream temperature control process, it is not explicitly represented in many stream temperature models, an omission that may reduce predictive accuracy, hinder management of aquatic habitat, and decrease user confidence. We assessed the performance of a previously-described process-guided deep learning model of stream temperature in the Delaware River Basin (USA). We found lower accuracy (root mean square error [RMSE] of 1.71 versus 1.35°C) and stronger seasonal bias (absolute mean monthly bias of 1.06 vs. 0.68°C) for reaches primarily influenced by deep groundwater as compared to atmospheric conditions. We then tested four approaches for improving groundwater process representation: (a) a custom loss function leveraging the unique patterns of air and water temperature coupling characteristic of different temperature drivers, (b) inclusion of additional groundwater-relevant catchment attributes, (c) incorporation of additional process model outputs, and (d) a composite model. The custom loss function and the additional attributes significantly improved the predictive accuracy in groundwater-dominated reaches (RMSE of 1.37 and 1.26°C) and reduced the seasonal bias (absolute mean monthly bias of 0.44 and 0.48°C), but neither approach could identify holdout groundwater reaches. Variable importance analysis indicates the custom loss function nudges the model to use the existing inputs more efficiently, whereas with the added features the model relies on a broader suite of inputs. This analysis is a substantial step toward more accurately representing groundwater discharge processes in stream temperature models and will improve predictive accuracy and inform habitat management.

54 ENVIRONMENTAL SCIENCES↗

Investigation of TRPV1 loss-of-function phenotypes in TRPV1 Leu206Stop mice generated by N-ethyl-N-nitrosourea mutagenesis

Highlights: • N-ethyl-N-nitrosourea mutagenesis can generate mouse models for pain research. • TRPV1 Leu206Stop mice closely resemble the phenotype of TRPV1 knockout mice. • Mechanical hypersensitivity is not impaired in TRPV1 Leu206Stop mice. N-ethyl-N-nitrosourea (ENU) random mutagenesis was used to generate a mouse model for the analysis of the transient receptor potential vanilloid 1 (TRPV1) cation channel. A transversion from T→A in exon 4 led to a Leu206Stop mutation generating a loss-of-function mutant. The TRPV1 agonist capsaicin was used to analyze functional and nociceptive parameters in vitro and in vivo in TRPV1 Leu206Stop mice and congenic C3HeB/FeJ controls. Capsaicin-induced [Ca{sup 2+}]{sub i} changes in small diameter DRG neurons were significantly diminished in TRPV1 Leu206Stop mice and administration of capsaicin induced neither hypothermia nor nocifensive behaviour in vivo. TRPV1 Leu206Stop mice were tested in the spinal nerve ligation of mononeuropathic pain and developed mechanical hypersensitivity two weeks after nerve injury. In the open field test, a significant increase in spontaneous locomotion was detected in TRPV1 Leu206Stop mice as compared to wildtype controls. TRPV1 knockout mice have been reported to carry a similar phenotype regarding capsaicin-evoked responses in vitro and in vivo. However, in contrast to TRPV1 Leu206Stop mice, TRPV1 knockout mice did not differ in spontaneous locomotion as compared to congenic C57BL/6 mice, suggesting subtle ENU-dependent or independent strain differences between TRPV1 Leu206Stop mice and their wildtype controls. In summary, these data revealed a target-related (i.e. capsaicin-evoked) phenotype of TRPV1 Leu206Stop mice closely resembling that of published TRPV1 knockout mice. However, since ENU-mutant mice are congenic with the mouse strain initially used in random mutagenesis, direct phenotypic comparison with the respective wildtype controls is possible, and the time-consuming backcrossing in lines with targeted mutations is avoided.

60 APPLIED LIFE SCIENCES↗

Transcription Factor 4 loss-of-function is associated with deficits in progenitor proliferation and cortical neuron content

Transcription Factor 4 ( TCF4) has been associated with autism, schizophrenia, and other neuropsychiatric disorders. However, how pathological TCF4 mutations affect the human neural tissue is poorly understood. Here, we derive neural progenitor cells, neurons, and brain organoids from skin fibroblasts obtained from children with Pitt-Hopkins Syndrome carrying clinically relevant mutations in TCF4 . We show that neural progenitors bearing these mutations have reduced proliferation and impaired capacity to differentiate into neurons. We identify a mechanism through which TCF4 loss-of-function leads to decreased Wnt signaling and then to diminished expression of SOX genes, culminating in reduced progenitor proliferation in vitro. Moreover, we show reduced cortical neuron content and impaired electrical activity in the patient-derived organoids, phenotypes that were rescued after correction of TCF4 expression or by pharmacological modulation of Wnt signaling. This work delineates pathological mechanisms in neural cells harboring TCF4 mutations and provides a potential target for therapeutic strategies for genetic disorders associated with this gene.

59 BASIC BIOLOGICAL SCIENCES↗

Structures of PHR Domains from Mus musculus Phr1 (Mycbp2) Explain the Loss-of-Function Mutation (Gly1092 → Glu) of the C. elegans Ortholog RPM-1

PHR [PAM (protein associated with Myc)-HIW (Highwire)-RPM-1 (regulator of presynaptic morphology 1)] proteins are conserved, large multi-domain E3 ubiquitin ligases with modular architecture. PHR proteins presynaptically control synaptic growth and axon guidance and postsynaptically regulate endocytosis of glutamate receptors. Dysfunction of neuronal ubiquitin-mediated proteasomal degradation is implicated in various neurodegenerative diseases. PHR proteins are characterized by the presence of two PHR domains near the N-terminus, which are essential for proper localization and function. Structures of both the first and second PHR domains of Mus musculus (mouse) Phr1 (MYC binding protein 2, Mycbp2) have been determined, revealing a novel {beta} sandwich fold composed of 11 antiparallel {beta}-strands. Conserved loops decorate the apical side of the first PHR domain (MmPHR1), yielding a distinct conserved surface feature. The surface of the second PHR domain (MmPHR2), in contrast, lacks significant conservation. Importantly, the structure of MmPHR1 provides insights into a loss-of-function mutation, Gly1092 {yields} Glu, observed in the Caenorhabditis elegans ortholog RPM-1.

59 BASIC BIOLOGICAL SCIENCES↗

LossLens: Diagnostics for Machine Learning Through Loss Landscape Visual Analytics

Modern machine learning often relies on optimizing a neural network's parameters using a loss function to learn complex features. Beyond training, examining the loss function with respect to a network's parameters (i.e., as a loss landscape) can reveal insights into the architecture and learning process. While the local structure of the loss landscape surrounding an individual solution can be characterized using a variety of approaches, the global structure of a loss landscape, which includes potentially many local minima corresponding to different solutions, remains far more difficult to conceptualize and visualize. To address this difficulty, we introduce LossLens, a visual analytics framework that explores loss landscapes at multiple scales. LossLens integrates metrics from global and local scales into a comprehensive visual representation, enhancing model diagnostics. Here we demonstrate LossLens through two case studies: visualizing how residual connections influence a ResNet-20, and visualizing how physical parameters influence a physics-informed neural network (PINN) solving a simple convection problem.

97 MATHEMATICS AND COMPUTING↗

Star formation through thermal instability of radiative plasma with finite electron inertia and finite Larmor radius corrections

I have studied the effects of finite electron inertia, finite ion Larmor radius (FLR) corrections, and radiative heat-loss function on the thermal instability of an infinite homogeneous, viscous plasma incorporating the effect of thermal conductivity for star formation in interstellar medium (ISM). A general dispersion relation is derived using the normal mode analysis method with the help of relevant linearized perturbation equations of the problem. The wave propagation is discussed for longitudinal and transverse directions to the external magnetic field and the conditions of modified thermal instabilities and stabilities are discussed in different cases. We find that the thermal instability criterion is get modified into radiative instability criterion by inclusion of radiative heat-loss functions with thermal conductivity. The viscosity of medium removes the effect of FLR corrections from the condition of radiative instability. Numerical calculation shows stabilizing effect of heat-loss function, viscosity and FLR corrections, and destabilizing effect of finite electron inertia on the thermal instability. Results carried out in this paper shows that stars are formed in interstellar medium mainly due to thermal instability.

79 ASTRONOMY AND ASTROPHYSICS↗

A strong loss-of-function mutation in RAN1 results in constitutive activation of the ethylene response pathway as well as a rosette-lethal phenotype

A recessive mutation was identified that constitutively activated the ethylene response pathway in Arabidopsis and resulted in a rosette-lethal phenotype. Positional cloning of the gene corresponding to this mutation revealed that it was allelic to responsive to antagonist1 (ran1), a mutation that causes seedlings to respond in a positive manner to what is normally a competitive inhibitor of ethylene binding. In contrast to the previously identified ran1-1 and ran1-2 alleles that are morphologically indistinguishable from wild-type plants, this ran1-3 allele results in a rosette-lethal phenotype. The predicted protein encoded by the RAN1 gene is similar to the Wilson and Menkes disease proteins and yeast Ccc2 protein, which are integral membrane cation-transporting P-type ATPases involved in copper trafficking. Genetic epistasis analysis indicated that RAN1 acts upstream of mutations in the ethylene receptor gene family. However, the rosette-lethal phenotype of ran1-3 was not suppressed by ethylene-insensitive mutants, suggesting that this mutation also affects a non-ethylene-dependent pathway regulating cell expansion. The phenotype of ran1-3 mutants is similar to loss-of-function ethylene receptor mutants, suggesting that RAN1 may be required to form functional ethylene receptors. Furthermore, these results suggest that copper is required not only for ethylene binding but also for the signaling function of the ethylene receptors.

NASA Discipline Plant Biology↗

Cooling of solar flares plasmas. 1: Theoretical considerations

Theoretical models of the cooling of flare plasma are reexamined. By assuming that the cooling occurs in two separate phase where conduction and radiation, respectively, dominate, a simple analytic formula for the cooling time of a flare plasma is derived. Unlike earlier order-of-magnitude scalings, this result accounts for the effect of the evolution of the loop plasma parameters on the cooling time. When the conductive cooling leads to an 'evaporation' of chromospheric material, the cooling time scales L(exp 5/6)/p(exp 1/6), where the coronal phase (defined as the time maximum temperature). When the conductive cooling is static, the cooling time scales as L(exp 3/4)n(exp 1/4). In deriving these results, use was made of an important scaling law (T proportional to n(exp 2)) during the radiative cooling phase that was forst noted in one-dimensional hydrodynamic numerical simulations (Serio et al. 1991; Jakimiec et al. 1992). Our own simulations show that this result is restricted to approximately the radiative loss function of Rosner, Tucker, & Vaiana (1978). for different radiative loss functions, other scaling result, with T and n scaling almost linearly when the radiative loss falls off as T(exp -2). It is shown that these scaling laws are part of a class of analytic solutions developed by Antiocos (1980).

Cargill, Peter J.↗

PPPC 4 DM secondary: a Poor Particle Physicist Cookbook for secondary radiation from Dark Matter

We enlarge the set of recipes and ingredients at disposal of any poor particle physicist eager to cook up signatures from weak-scale Dark Matter models by computing two secondary emissions due to DM particles annihilating or decaying in the galactic halo, namely the radio signals from synchrotron emission and the gamma rays from bremsstrahlung. We consider several magnetic field configurations and propagation scenarios for electrons and positrons. We also provide an improved energy loss function for electrons and positrons in the Galaxy, including synchrotron losses in the different configurations, bremsstrahlung losses, ionization losses and Inverse Compton losses with an updated InterStellar Radiation Field.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Real-Time Attitude Independent Three Axis Magnetometer Calibration

In this paper new real-time approaches for three-axis magnetometer sensor calibration are derived. These approaches rely on a conversion of the magnetometer-body and geomagnetic-reference vectors into an attitude independent observation by using scalar checking. The goal of the full calibration problem involves the determination of the magnetometer bias vector, scale factors and non-orthogonality corrections. Although the actual solution to this full calibration problem involves the minimization of a quartic loss function, the problem can be converted into a quadratic loss function by a centering approximation. This leads to a simple batch linear least squares solution. In this paper we develop alternative real-time algorithms based on both the extended Kalman filter and Unscented filter. With these real-time algorithms, a full magnetometer calibration can now be performed on-orbit during typical spacecraft mission-mode operations. Simulation results indicate that both algorithms provide accurate integer resolution in real time, but the Unscented filter is more robust to large initial condition errors than the extended Kalman filter. The algorithms are also tested using actual data from the Transition Region and Coronal Explorer (TRACE).

Crassidis, John L.↗

Generalized Canonical Polyadic Tensor Decomposition

Tensor decomposition is a fundamental unsupervised machine learning method in data science, with applications including network analysis and sensor data processing. This work develops a generalized canonical polyadic (GCP) low-rank tensor decomposition that allows other loss functions besides squared error. For instance, we can use logistic loss or Kullback--Leibler divergence, enabling tensor decomposition for binary or count data. We present a variety of statistically motivated loss functions for various scenarios. We provide a generalized framework for computing gradients and handling missing data that enables the use of standard optimization methods for fitting the model. Furthermore, we demonstrate the flexibility of the GCP decomposition on several real-world examples including interactions in a social network, neural activity in a mouse, and monthly rainfall measurements in India.

97 MATHEMATICS AND COMPUTING↗

Error minimizing algorithms for nearest eighbor classifiers

Stack Filters define a large class of discrete nonlinear filter first introd uced in image and signal processing for noise removal. In recent years we have suggested their application to classification problems, and investigated their relationship to other types of discrete classifiers such as Decision Trees. In this paper we focus on a continuous domain version of Stack Filter Classifiers which we call Ordered Hypothesis Machines (OHM), and investigate their relationship to Nearest Neighbor classifiers. We show that OHM classifiers provide a novel framework in which to train Nearest Neighbor type classifiers by minimizing empirical error based loss functions. We use the framework to investigate a new cost sensitive loss function that allows us to train a Nearest Neighbor type classifier for low false alarm rate applications. We report results on both synthetic data and real-world image data.

97 MATHEMATICS AND COMPUTING↗

RNA-seq latent featurizer using center loss cost function

Predictive modeling of patient tumor growth response to drug treatment is severely limited by a lack of experimental data for training. Combining experimental data from several studies is an attractive approach, but presents two problems: batch effects and limited data for individual drugs and cell lines. Batch effects are caused by systematic procedural differences among studies, which causes systematic experimental outcome differences. Directly using these experimental results as features for machine learning commonly causes problems when training on one study and testing on another. This severely limits a model's ability to perform well on new experiments. Even after combining studies, predicting outcomes on new patient tumors remains an open challenge. We propose a semi-supervised, autoencoder-based, machine learning procedure, which learns a smaller set of gene expression features that are robust to batch effects using background information on a cell line or tissue's tumor type. We implemented this reduced feature representation and show that the new feature space clusters strongly according to tumor type. This experiment is carried out across multiple studies: CCLE, CTRP, gCSI, GDSC, NCI60, and patient derived tumors. We hypothesize that using a batch effect resistant feature set across studies will improve prediction performance. Genome Data Commons (GDC) gene expression profiles for publicly available human tissues and cell lines from NCI60 and CCLE were processed using the semi-supervised learning procedure. Our autoencoder repurposes the 'center loss' (CL) cost function of Wen et. al. to learn a more generalized set of features using the cell line or tissue's tumor type.

He, Stewart↗

Addressing bias in bagging and boosting regression models

As artificial intelligence (AI) becomes widespread, there is increasing attention on investigating bias in machine learning (ML) models. Previous research concentrated on classification problems, with little emphasis on regression models. This paper presents an easy-to-apply and effective methodology for mitigating bias in bagging and boosting regression models, that is also applicable to any model trained through minimizing a differentiable loss function. Our methodology measures bias rigorously and extends the ML model's loss function with a regularization term to penalize high correlations between model errors and protected attributes. We applied our approach to three popular tree-based ensemble models: a random forest model (RF), a gradient-boosted model (GBT), and an extreme gradient boosting model (XGBoost). We implemented our methodology on a case study for predicting road-level traffic volume, where RF, GBT, and XGBoost models were shown to have high accuracy. Despite high accuracy, the ML models were shown to perform poorly on roads in minority-populated areas. Our bias mitigation approach reduced minority-related bias by over 50%.

97 MATHEMATICS AND COMPUTING↗