Extrapolation -- the ability to make inferences that go beyond the scope of one's experiences -- is a hallmark of human intelligence. By contrast, the generalization exhibited by contemporary neural network algorithms is largely limited to interpolation between data points in their training corpora. In this paper, we consider the challenge of learning representations that support extrapolation. We introduce a novel visual analogy benchmark that allows the graded evaluation of extrapolation as a function of distance from the convex domain defined by the training data. We also introduce a simple technique, context normalization, that encourages representations that emphasize the relations between objects. We find that this technique enables a significant improvement in the ability to extrapolate, considerably outperforming a number of competitive techniques.
Yang, Yuan; Pan, Ming; Beck, Hylke; Fisher, Colby; Beighley, R. Edward; Kao, Shih-Chieh; Hong, Yang; Wood, Eric
Abstract:
Conventional basin-by-basin approaches to calibrate hydrologic models are limited to gauged basins and typically result in spatially discontinuous parameter fields. Moreover, the consequent low calibration density in space falls seriously behind the need from present-day applications like high resolution river hydrodynamic modeling. In this study we calibrated three key parameters of the Variable Infiltration Capacity (VIC) model at every 1/8° grid-cell using machine learning-based maps of four streamflow characteristics for the conterminous United States (CONUS), with a total of 52,663 grid-cells. This new calibration approach, as an alternative to parameter regionalization, applied to ungauged regions too. A key difference made here is that we tried to regionalize physical variables (streamflow characteristics) instead of model parameters whose behavior may often be less well understood. The resulting parameter fields no longer presented any spatial discontinuities and the patterns corresponded well with climate characteristics, such as aridity and runoff ratio. The calibrated parameters were evaluated against observed streamflow from 704/648 (calibration/validation period) small-to-medium-sized catchments used to derive the streamflow characteristics, 3941/3809 (calibration/validation period) small-to-medium-sized catchments not used to derive the streamflow characteristics) as well as five large basins. Comparisons indicated marked improvements in bias and Nash-Sutcliffe efficiency. Model performance was still poor in arid and semiarid regions, which is mostly due to both model structural and forcing deficiencies. Although the performance gain was limited by the relative small number of parameters to calibrate, the study and results here served as a proof-of-concept for a new promising approach for fine-scale hydrologic model calibrations.
Taylor, Jenny A.; Bratton, Benjamin P.; Sichel, Sophie R.; Blair, Kris M.; Jacobs, Holly M.; DeMeester, Kristen E.; Kuru, Erkin; Gray, Joe; Biboy, Jacob; VanNieuwenhze, Michael S.; Vollmer, Waldemar; Grimes, Catherine L.; Shaevitz, Joshua W.; Salama, Nina R.
Abstract:
Helical cell shape is necessary for efficient stomach colonization by Helicobacter pylori, but the molecular mechanisms for generating helical shape remain unclear. We show that the helical centerline pitch and radius of wild-type H. pylori cells dictate surface curvatures of considerably higher positive and negative Gaussian curvatures than those present in straight- or curved-rod bacteria. Quantitative 3D microscopy analysis of short pulses with either N-acetylmuramic acid or D-alanine metabolic probes showed that cell wall growth is enhanced at both sidewall curvature extremes. Immunofluorescence revealed MreB is most abundant at negative Gaussian curvature, while the bactofilin CcmA is most abundant at positive Gaussian curvature. Strains expressing CcmA variants with altered polymerization properties lose helical shape and associated positive Gaussian curvatures. We thus propose a model where CcmA and MreB promote PG synthesis at positive and negative Gaussian curvatures, respectively, and that this patterning is one mechanism necessary for maintaining helical shape.
Since 1850 the concentration of atmospheric methane (CH4), a potent greenhouse gas, has more than doubled. Recent studies suggest that emission inventories may be missing sources and underestimating emissions. To investigate whether offshore oil and gas platforms leak CH4 during normal operation, we measured CH4 mole fractions around eight oil and gas production platforms in the North Sea which were neither flaring gas nor off-loading oil. We use the measurements from summer 2017, along with meteorological data, in a Gaussian plume model to estimate CH4 emissions from each platform. We find CH4 mole fractions of between 11 and 370 ppb above background concentrations downwind of the platforms measured, corresponding to a median CH4 emission of 6.8 g CH4 s-1 for each platform, with a range of 2.9 to 22.3 g CH4 s-1. When matched to production records, during our measurements individual platforms lost between 0.04% and 1.4% of gas produced with a median loss of 0.23%. When the measured platforms are considered collectively, (i.e. the sum of platforms’ emission fluxes weighted by the sum of the platforms’ production), we estimate the CH4 loss to be 0.19% of gas production. These estimates are substantially higher than the emissions most recently reported to the National Atmospheric Emission Inventory (NAEI) for total CH4 loss from United Kingdom platforms in the North Sea. The NAEI reports CH4 losses from the offshore oil and gas platforms we measured to be 0.13% of gas production, with most of their emissions coming from gas flaring and offshore oil loading, neither of which were taking place at the time of our measurements. All oil and gas platforms we observed were found to leak CH4 during normal operation and much of this leakage has not been included in UK emission inventories. Further research is required to accurately determine total CH4 leakage from all offshore oil and gas operations and to properly include the leakage in national and international emission inventories.
