scholarly journals Decreasing wind speed extrapolation error via domain-specific feature extraction and selection

2020 ◽  
Vol 5 (3) ◽  
pp. 959-975
Author(s):  
Daniel Vassallo ◽  
Raghavendra Krishnamurthy ◽  
Harindra J. S. Fernando
Keyword(s):  
Wind Speed ◽  
Power Law ◽  
Domain Knowledge ◽  
Test Case ◽  
Domain Specific ◽  
Promising Tool ◽  
Feature Extraction And Selection ◽  
Log Law ◽  
Wind Resources

Abstract. Model uncertainty is a significant challenge in the wind energy industry and can lead to mischaracterization of millions of dollars' worth of wind resources. Machine learning methods, notably deep artificial neural networks (ANNs), are capable of modeling turbulent and chaotic systems and offer a promising tool to produce high-accuracy wind speed forecasts and extrapolations. This paper uses data collected by profiling Doppler lidars over three field campaigns to investigate the efficacy of using ANNs for wind speed vertical extrapolation in a variety of terrains, and it quantifies the role of domain knowledge in ANN extrapolation accuracy. A series of 11 meteorological parameters (features) are used as ANN inputs, and the resulting output accuracy is compared with that of both standard log-law and power-law extrapolations. It is found that extracted nondimensional inputs, namely turbulence intensity, current wind speed, and previous wind speed, are the features that most reliably improve the ANN's accuracy, providing up to a 65 % and 52 % increase in extrapolation accuracy over log-law and power-law predictions, respectively. The volume of input data is also deemed important for achieving robust results. One test case is analyzed in depth using dimensional and nondimensional features, showing that the feature nondimensionalization drastically improves network accuracy and robustness for sparsely sampled atmospheric cases.

scholarly journals Decreasing Wind Speed Extrapolation Error via Domain-Specific Feature Extraction and Selection

2019 ◽  
Author(s):  
Daniel Vassallo ◽  
Raghavendra Krishnamurthy ◽  
Harindra J. S. Fernando
Keyword(s):  
Wind Speed ◽  
Power Law ◽  
Domain Knowledge ◽  
Test Case ◽  
Domain Specific ◽  
Promising Tool ◽  
Feature Extraction And Selection ◽  
Using Data ◽  
Meteorological Features ◽  
Wind Resource

Abstract. Model uncertainty is a significant challenge in the wind energy industry and can lead to mischaracterization of millions of dollars' worth of wind resource. Machine learning methods, notably deep artificial neural networks (ANNs), are capable of modeling turbulent and chaotic systems and offer a promising tool that can be used to produce high-accuracy wind speed forecasts and extrapolations. This paper quantifies the role of domain knowledge on ANN wind speed extrapolation accuracy using data collected using profiling lidars over three field campaigns. A series of 11 meteorological features are used as ANN inputs and the resulting output accuracy is compared with that of a simple power law extrapolation. It is found that normalized inputs, namely turbulence intensity, normalized current wind speed, and normalized previous wind speed, are the features that most reliably improve ANN accuracy, providing up to a 52 % increase in extrapolation accuracy over the power law predictions. The volume of input data is also deemed important for achieving robust results. One test case is analyzed in-depth using dimensional and non-dimensional features, showing that feature normalization drastically improves network accuracy and robustness for uncommon atmospheric cases.

Testing the sociometer: The role of self-esteem as a domain specific mediator

2007 ◽  
Author(s):  
P. S. Kavanagh ◽  
G. J. O. Fletcher ◽  
B. J. Ellis
Keyword(s):  
Self Esteem ◽  
Domain Specific

When Do Auditors Use Specialists' Work to Improve Problem Representations of and Judgments about Complex Estimates?

2017 ◽  
Vol 93 (4) ◽  
pp. 177-202 ◽  
Author(s):  
Emily E. Griffith
Keyword(s):  
Pattern Recognition ◽  
Situational Factors ◽  
Situational Factor ◽  
High Quality ◽  
Quality Audit ◽  
Domain Specific ◽  
Epistemic Motivation ◽  
Specific Expertise ◽  
Problem Representations

ABSTRACT Auditors are more likely to identify misstatements in complex estimates if they recognize problematic patterns among an estimate's underlying assumptions. Rich problem representations aid pattern recognition, but auditors likely have difficulty developing them given auditors' limited domain-specific expertise in this area. In two experiments, I predict and find that a relational cue in a specialist's work highlighting aggressive assumptions improves auditors' problem representations and subsequent judgments about estimates. However, this improvement only occurs when a situational factor (e.g., risk) increases auditors' epistemic motivation to incorporate the cue into their problem representations. These results suggest that auditors do not always respond to cues in specialists' work. More generally, this study highlights the role of situational factors in increasing auditors' epistemic motivation to develop rich problem representations, which contribute to high-quality audit judgments in this and other domains where pattern recognition is important.

scholarly journals Resource and dependency based test case generation for RESTful Web services

2021 ◽  
Vol 26 (4) ◽  
Author(s):  
Man Zhang ◽  
Bogdan Marculescu ◽  
Andrea Arcuri
Keyword(s):  
Web Services ◽  
Domain Knowledge ◽  
Automated System ◽  
Test Case ◽  
Test Cases ◽  
Evolutionary Search ◽  
Sampling Strategies ◽  
Mutation Operators ◽  
On Line ◽  
Restful Web Services

AbstractNowadays, RESTful web services are widely used for building enterprise applications. REST is not a protocol, but rather it defines a set of guidelines on how to design APIs to access and manipulate resources using HTTP over a network. In this paper, we propose an enhanced search-based method for automated system test generation for RESTful web services, by exploiting domain knowledge on the handling of HTTP resources. The proposed techniques use domain knowledge specific to RESTful web services and a set of effective templates to structure test actions (i.e., ordered sequences of HTTP calls) within an individual in the evolutionary search. The action templates are developed based on the semantics of HTTP methods and are used to manipulate the web services’ resources. In addition, we propose five novel sampling strategies with four sampling methods (i.e., resource-based sampling) for the test cases that can use one or more of these templates. The strategies are further supported with a set of new, specialized mutation operators (i.e., resource-based mutation) in the evolutionary search that take into account the use of these resources in the generated test cases. Moreover, we propose a novel dependency handling to detect possible dependencies among the resources in the tested applications. The resource-based sampling and mutations are then enhanced by exploiting the information of these detected dependencies. To evaluate our approach, we implemented it as an extension to the EvoMaster tool, and conducted an empirical study with two selected baselines on 7 open-source and 12 synthetic RESTful web services. Results show that our novel resource-based approach with dependency handling obtains a significant improvement in performance over the baselines, e.g., up to + 130.7% relative improvement (growing from + 27.9% to + 64.3%) on line coverage.

scholarly journals Templated Text Synthesis for Expert-Guided Multi-Label Extraction from Radiology Reports

2021 ◽  
Vol 3 (2) ◽  
pp. 299-317
Author(s):  
Patrick Schrempf ◽  
Hannah Watson ◽  
Eunsoo Park ◽  
Maciej Pajak ◽  
Hamish MacKinnon ◽  
...  
Keyword(s):  
Domain Knowledge ◽  
Expert Knowledge ◽  
Medical Image Analysis ◽  
Imaging Data ◽  
Data Synthesis ◽  
Domain Specific ◽  
Radiology Reports ◽  
Training Examples ◽  
Analysis Models ◽  
Difficult Cases

Training medical image analysis models traditionally requires large amounts of expertly annotated imaging data which is time-consuming and expensive to obtain. One solution is to automatically extract scan-level labels from radiology reports. Previously, we showed that, by extending BERT with a per-label attention mechanism, we can train a single model to perform automatic extraction of many labels in parallel. However, if we rely on pure data-driven learning, the model sometimes fails to learn critical features or learns the correct answer via simplistic heuristics (e.g., that “likely” indicates positivity), and thus fails to generalise to rarer cases which have not been learned or where the heuristics break down (e.g., “likely represents prominent VR space or lacunar infarct” which indicates uncertainty over two differential diagnoses). In this work, we propose template creation for data synthesis, which enables us to inject expert knowledge about unseen entities from medical ontologies, and to teach the model rules on how to label difficult cases, by producing relevant training examples. Using this technique alongside domain-specific pre-training for our underlying BERT architecture i.e., PubMedBERT, we improve F1 micro from 0.903 to 0.939 and F1 macro from 0.512 to 0.737 on an independent test set for 33 labels in head CT reports for stroke patients. Our methodology offers a practical way to combine domain knowledge with machine learning for text classification tasks.

scholarly journals Total pancreatectomy and pancreatic fistula: friend or foe?

2021 ◽  
Author(s):  
Roberto Salvia ◽  
Gabriella Lionetto ◽  
Giampaolo Perri ◽  
Giuseppe Malleo ◽  
Giovanni Marchegiani
Keyword(s):  
High Risk ◽  
Pancreatic Fistula ◽  
Total Pancreatectomy ◽  
Postoperative Pancreatic Fistula ◽  
Current Evidence ◽  
Perioperative Outcomes ◽  
Surgical Morbidity ◽  
Pancreatic Anastomosis ◽  
Promising Tool

AbstractPostoperative pancreatic fistula (POPF) still represents the major driver of surgical morbidity after pancreaticoduodenectomy. The purpose of this narrative review was to critically analyze current evidence supporting the use of total pancreatectomy (TP) to prevent the development of POPF in patients with high-risk pancreas, and to explore the role of completion total pancreatectomy (CP) in the management of severe POPF. Considering the encouraging perioperative outcomes, TP may represent a promising tool to avoid the morbidity related to an extremely high-risk pancreatic anastomosis in selected patients. Surgical management of severe POPF is only required in few critical scenarios. In this context, even if anecdotal, CP might play a role as last resort in expert hands.

scholarly journals Toward a Better Understanding of Metal Nanoparticles, a Novel Strategy from Eucalyptus Plants

Plants ◽  
2021 ◽  
Vol 10 (5) ◽  
pp. 929
Author(s):  
Hanadi Sawalha ◽  
Rambod Abiri ◽  
Ruzana Sanusi ◽  
Noor Azmi Shaharuddin ◽  
Aida Atiqah Mohd Noor ◽  
...  
Keyword(s):  
Metal Nanoparticles ◽  
Woody Plants ◽  
Woody Species ◽  
Forest Products ◽  
Ideal Candidate ◽  
Promising Tool ◽  
Novel Strategy ◽  
The Ideal

Nanotechnology is a promising tool that has opened the doors of improvement to the quality of human’s lives through its potential in numerous technological aspects. Green chemistry of nanoscale materials (1–100 nm) is as an effective and sustainable strategy to manufacture homogeneous nanoparticles (NPs) with unique properties, thus making the synthesis of green NPs, especially metal nanoparticles (MNPs), the scientist’s core theme. Researchers have tested different organisms to manufacture MNPs and the results of experiments confirmed that plants tend to be the ideal candidate amongst all entities and are suitable to synthesize a wide variety of MNPs. Natural and cultivated Eucalyptus forests are among woody plants used for landscape beautification and as forest products. The present review has been written to reflect the efficacious role of Eucalyptus in the synthesis of MNPs. To better understand this, the route of extracting MNPs from plants, in general, and Eucalyptus, in particular, are discussed. Furthermore, the crucial factors influencing the process of MNP synthesis from Eucalyptus as well as their characterization and recent applications are highlighted. Information gathered in this review is useful to build a basis for new prospective research ideas on how to exploit this woody species in the production of MNPs. Nevertheless, there is a necessity to feed the scientific field with further investigations on wider applications of Eucalyptus-derived MNPs.

scholarly journals Dust Production around Carbon-Rich Stars: The Role of Metallicity

Universe ◽  
2021 ◽  
Vol 7 (7) ◽  
pp. 233
Author(s):  
Ambra Nanni ◽  
Sergio Cristallo ◽  
Jacco Th. van Loon ◽  
Martin A. T. Groenewegen
Keyword(s):  
Wind Speed ◽  
Galactic Halo ◽  
Magellanic Clouds ◽  
Asymptotic Giant Branch ◽  
Agb Stars ◽  
Circumstellar Envelopes ◽  
Chemical Enrichment ◽  
Wide Range ◽  
The Universe

Background: Most of the stars in the Universe will end their evolution by losing their envelope during the thermally pulsing asymptotic giant branch (TP-AGB) phase, enriching the interstellar medium of galaxies with heavy elements, partially condensed into dust grains formed in their extended circumstellar envelopes. Among these stars, carbon-rich TP-AGB stars (C-stars) are particularly relevant for the chemical enrichment of galaxies. We here investigated the role of the metallicity in the dust formation process from a theoretical viewpoint. Methods: We coupled an up-to-date description of dust growth and dust-driven wind, which included the time-averaged effect of shocks, with FRUITY stellar evolutionary tracks. We compared our predictions with observations of C-stars in our Galaxy, in the Magellanic Clouds (LMC and SMC) and in the Galactic Halo, characterised by metallicity between solar and 1/10 of solar. Results: Our models explained the variation of the gas and dust content around C-stars derived from the IRS Spitzer spectra. The wind speed of the C-stars at varying metallicity was well reproduced by our description. We predicted the wind speed at metallicity down to 1/10 of solar in a wide range of mass-loss rates.

Sign in / Sign up

Export Citation Format

Share Document