•Optimal bidding strategies for reservoir hydro with supervised learning.•Choosing stochastic or deterministic bidding for day ahead electricity production.•Integrating machine learning with existing ...model framework.•Improving profits and reducing calculation time.
Power producers use a wide range of decision support systems to manage and plan for sales in the day-ahead electricity market. The available tools have advantages and disadvantages and the operators are often faced with the challenge of choosing the most advantageous bidding strategy for any given day. Since only one bid can be submitted each day, this choice can not be avoided. The optimal solution is not known until after spot clearing. Results from the models and strategy used, and their impact on profitability, can either be continuously registered, or simulated with use of historic data. Access to an increasing amount of data opens for the application of machine learning models to predict the best combination of models and strategy for any given day. In this article, historical performance of two given bidding strategies over several years have been analyzed with a combination of domain knowledge and machine learning techniques. A wide range of model variables accessible prior to bidding have been evaluated to predict the optimal strategy for a given day. Results indicate that a machine learning model can learn to slightly outperform a static strategy where one bidding method is chosen based on overall historic performance.
The absolute neutrino mass scale is currently unknown, but can be constrained by cosmology. The WiggleZ high redshift, star-forming, and blue galaxy sample offers a complementary data set to previous ...surveys for performing these measurements, with potentially different systematics from nonlinear structure formation, redshift-space distortions, and galaxy bias. We obtain a limit of capital sigma m sub(nu) < 0.60 eV (95% confidence) for WiggleZ + Wilkinson Microwave Anisotropy Probe. Combining with priors on the Hubble parameter and the baryon acoustic oscillation scale gives capital sigma m sub(nu) < 0.29 eV, which is the strongest neutrino mass constraint derived from spectroscopic galaxy redshift surveys.
Despite its continued observational successes, there is a persistent (and growing) interest in extending cosmology beyond the standard model, ΛCDM. This is motivated by a range of apparently serious ...theoretical issues, involving such questions as the cosmological constant problem, the particle nature of dark matter, the validity of general relativity on large scales, the existence of anomalies in the CMB and on small scales, and the predictivity and testability of the inflationary paradigm. In this paper, we summarize the current status of ΛCDM as a physical theory, and review investigations into possible alternatives along a number of different lines, with a particular focus on highlighting the most promising directions. While the fundamental problems are proving reluctant to yield, the study of alternative cosmologies has led to considerable progress, with much more to come if hopes about forthcoming high-precision observations and new theoretical ideas are fulfilled.
The identification of gene-gene and gene-environment interactions in genome-wide association studies is challenging due to the unknown nature of the interactions and the overwhelmingly large number ...of possible combinations. Parametric regression models are suitable to look for prespecified interactions. Nonparametric models such as tree ensemble models, with the ability to detect any unspecified interaction, have previously been difficult to interpret. However, with the development of methods for model explainability, it is now possible to interpret tree ensemble models efficiently and with a strong theoretical basis.
We propose a tree ensemble- and SHAP-based method for identifying as well as interpreting potential gene-gene and gene-environment interactions on large-scale biobank data. A set of independent cross-validation runs are used to implicitly investigate the whole genome. We apply and evaluate the method using data from the UK Biobank with obesity as the phenotype. The results are in line with previous research on obesity as we identify top SNPs previously associated with obesity. We further demonstrate how to interpret and visualize interaction candidates.
The new method identifies interaction candidates otherwise not detected with parametric regression models. However, further research is needed to evaluate the uncertainties of these candidates. The method can be applied to large-scale biobanks with high-dimensional data.
Celotno besedilo
Dostopno za:
DOBA, IZUM, KILJ, NUK, PILJ, PNG, SAZU, SIK, UILJ, UKNU, UL, UM, UPUK
Estimating feature importance, which is the contribution of a prediction or several predictions due to a feature, is an essential aspect of explaining data-based models. Besides explaining the model ...itself, an equally relevant question is which features are important in the underlying data generating process. We present a Shapley-value-based framework for inferring the importance of individual features, including uncertainty in the estimator. We build upon the recently published model-agnostic feature importance score of SAGE (Shapley additive global importance) and introduce Sub-SAGE. For tree-based models, it has the advantage that it can be estimated without computationally expensive resampling. We argue that for all model types the uncertainties in our Sub-SAGE estimator can be estimated using bootstrapping and demonstrate the approach for tree ensemble methods. The framework is exemplified on synthetic data as well as large genotype data for predicting feature importance with respect to obesity.
Celotno besedilo
Dostopno za:
DOBA, IZUM, KILJ, NUK, PILJ, PNG, SAZU, SIK, UILJ, UKNU, UL, UM, UPUK
Context. Recent findings of line emission at 3.5 keV in both individual and stacked X-ray spectra of galaxy clusters have been speculated to have dark matter origin. Aims. If the origin is indeed ...dark matter, the emission line is expected to be detectable from the Milky Way dark matter halo. Methods. We perform a line search in public Chandra X-ray observations of the region near Sgr A*. We derive upper limits on the line emission flux for the 2.0−9.0 keV energy interval and discuss their potential physical interpretations including various scenarios of decaying and annihilating dark matter. Results. While we find no clear evidence for its presence, the upper flux limits are not inconsistent with the recent detections for conservative mass profiles of the Milky Way. Conclusions. The results depend mildly on the spectral modelling, and strongly on the choice of dark matter profile.
The balancing market for power is designed to account for the difference between predicted supply/demand of electricity and the realised supply/demand. However, increased electrification of society ...changes the consumption patterns, and increased production from renewable sources leads to larger un-predicted fluctuations in production, both effects potentially leading to increased balancing. We analyse public market data for the balancing market (manual Frequency Restoration Reserve) for Norway from 2016 to 2022 to investigate and document these effects. The data is newer than for similar analyses and the eight years of data is more than double the time span previously covered.
The main findings are: (a) The balancing volumes are dominated by hours of zero regulation but for non-zero hours, the balancing volumes are increasing during the eight-year period. (b) The balancing prices are primarily correlated with day-ahead prices and secondary with balancing volumes. The latter correlation is found to be increasingly non-linear with time. (c) The balancing volumes and the price difference between balancing price and day-ahead price are strongly correlated with the previous hour. (d) The increasing share of wind power has not impacted the frequency of balancing, which has remained stable during the 8 years studied. However, the volumes and share of balancing power compared to overall production have increased, suggesting that the hours which are inherently difficult to predict remain the same. (e) Market data alone cannot predict balancing volumes. If attempting, the auto-correlation becomes the main source of information.
•Statistical analysis of the Norwegian regulated power market between 2016 and 2022.•Showcasing trends, changes and predictability.•Influence of wind power expansion on regulation volumes and regulation market dynamics.
We consider the possibility of constraining decaying dark matter by looking out through the Milky Way halo. Specifically, we use Chandra blank sky observations to constrain the parameter space of ...sterile neutrinos. We find that a broad band in parameter space is still open, leaving the sterile neutrino as an excellent dark matter candidate.
The electricity market is driven by complicated interactions that are hard to model analytically. This is particularly the case for the balancing market, where imbalances between supply and demand ...after the day-ahead market clearance are balanced. The balancing market bridges the gap between the day-ahead market and the actual power system operations. Being able to predict the necessary balancing volumes and prices some hours in advance of the operational hour will allow power producers to plan their production and trading in a more optimal way. There exist large amounts of open data that could contain predictive information about the balancing market, including day-ahead market data and climatic data. However, the literature on forecasting volume and prices in the balancing market is sparse compared to the rich literature on forecasting for the day-ahead market. Neural networks are powerful functional approximators and well-suited to model the complex relationships in the power market. It may also be used to study the predictability of the balancing volumes and prices forward in time. In this paper, we develop a model based on long short-term memory (LSTM) recurrent neural networks to predict volumes and prices in the Nordic balancing market based on public accessible data. Results show that the LSTM model performs well when compared to the two baselines selected. However, the performance is not significantly better, which indicates that the market data does not hold significant predictive information.
Two new high-precision measurements of the deuterium abundance from absorbers along the line of sight to the quasar PKS1937–1009 were presented. The absorbers have lower neutral hydrogen column ...densities (N(HI) ≈ 18 cm − 2 ) than for previous high-precision measurements, boding well for further extensions of the sample due to the plenitude of low column density absorbers. The total high-precision sample now consists of 12 measurements with a weighted average deuterium abundance of D/H = 2 . 55 ± 0 . 02 × 10 − 5 . The sample does not favour a dipole similar to the one detected for the fine structure constant. The increased precision also calls for improved nucleosynthesis predictions. For that purpose we have updated the public AlterBBN code including new reactions, updated nuclear reaction rates, and the possibility of adding new physics such as dark matter. The standard Big Bang Nucleosynthesis prediction of D/H = 2 . 456 ± 0 . 057 × 10 − 5 is consistent with the observed value within 1.7 standard deviations.