Matches in SemOpenAlex for { <https://semopenalex.org/work/W4379284259> ?p ?o ?g. }
Showing items 1 to 79 of
79
with 100 items per page.
- W4379284259 endingPage "120880" @default.
- W4379284259 startingPage "120880" @default.
- W4379284259 abstract "Uncorrelated flow images are used to study complex features occurring in phase change processes. More specifically, machine-learning algorithms relied on three experimental datasets to classify and quantify key flow parameters of R-134a condensing as it flows upwards in a vertical straight tube. Dataset-1 has 159 data points of vapor and liquid superficial flow velocities, which were correlated within four classical flow regimes (climbing film, flow reversal, churn, and slug). The data was used to train three machine-learning algorithms (Gaussian Naive Bayes, Support Vector Machines, and the k Nearest Neighbors) and to create continuous superficial flow velocities maps for flow conditions that were not necessarily tested experimentally showing good consistency. Dataset-2 generated a sequence of 2,453 uncorrelated flow images from 25 videos, which were pre-processed and used to train Decision Tree Classifier and Convolutional Neural Network algorithms to classify flow regimes based on distinct data splitting methods. The results showed that the splitting method used significantly impacts the algorithm’s accuracy. For instance, random splitting returned F-score values higher than 97%, while splitting by experiment presented much lower F-score values. Dataset-3 had 310,839 pre-processed flow images, which were correlated with heat transfer, quality, void fraction, and mass flow rate. These images were used to train a Convolutional Neural Network algorithm, which was able quantify the three first parameters presenting relative errors in the order 7%, 1% and 1%, respectively, and classify the mass flow rate with an accuracy of 99% considering a full window size. Finally, the Kolmogorov-Smirnov test was used to further explore the different accuracies obtained with two data splitting methods." @default.
- W4379284259 created "2023-06-05" @default.
- W4379284259 creator A5041426283 @default.
- W4379284259 creator A5060810519 @default.
- W4379284259 creator A5068091970 @default.
- W4379284259 date "2023-08-01" @default.
- W4379284259 modified "2023-10-16" @default.
- W4379284259 title "Multi-parameter classification and quantification of R-134a condensation using machine learning" @default.
- W4379284259 cites W2005096126 @default.
- W4379284259 cites W2013281957 @default.
- W4379284259 cites W2016903811 @default.
- W4379284259 cites W2133665775 @default.
- W4379284259 cites W2295107390 @default.
- W4379284259 cites W2326400787 @default.
- W4379284259 cites W2472739243 @default.
- W4379284259 cites W2531473284 @default.
- W4379284259 cites W2626492173 @default.
- W4379284259 cites W2803556351 @default.
- W4379284259 cites W2910653063 @default.
- W4379284259 cites W2914064718 @default.
- W4379284259 cites W2972468788 @default.
- W4379284259 cites W3009241579 @default.
- W4379284259 cites W3137714383 @default.
- W4379284259 cites W3186362729 @default.
- W4379284259 cites W3186396645 @default.
- W4379284259 cites W4220769844 @default.
- W4379284259 cites W4281710148 @default.
- W4379284259 cites W4367310569 @default.
- W4379284259 doi "https://doi.org/10.1016/j.applthermaleng.2023.120880" @default.
- W4379284259 hasPublicationYear "2023" @default.
- W4379284259 type Work @default.
- W4379284259 citedByCount "0" @default.
- W4379284259 crossrefType "journal-article" @default.
- W4379284259 hasAuthorship W4379284259A5041426283 @default.
- W4379284259 hasAuthorship W4379284259A5060810519 @default.
- W4379284259 hasAuthorship W4379284259A5068091970 @default.
- W4379284259 hasConcept C11413529 @default.
- W4379284259 hasConcept C119857082 @default.
- W4379284259 hasConcept C12267149 @default.
- W4379284259 hasConcept C153180895 @default.
- W4379284259 hasConcept C154945302 @default.
- W4379284259 hasConcept C175291020 @default.
- W4379284259 hasConcept C199360897 @default.
- W4379284259 hasConcept C33923547 @default.
- W4379284259 hasConcept C41008148 @default.
- W4379284259 hasConcept C50644808 @default.
- W4379284259 hasConcept C52001869 @default.
- W4379284259 hasConceptScore W4379284259C11413529 @default.
- W4379284259 hasConceptScore W4379284259C119857082 @default.
- W4379284259 hasConceptScore W4379284259C12267149 @default.
- W4379284259 hasConceptScore W4379284259C153180895 @default.
- W4379284259 hasConceptScore W4379284259C154945302 @default.
- W4379284259 hasConceptScore W4379284259C175291020 @default.
- W4379284259 hasConceptScore W4379284259C199360897 @default.
- W4379284259 hasConceptScore W4379284259C33923547 @default.
- W4379284259 hasConceptScore W4379284259C41008148 @default.
- W4379284259 hasConceptScore W4379284259C50644808 @default.
- W4379284259 hasConceptScore W4379284259C52001869 @default.
- W4379284259 hasFunder F4320322025 @default.
- W4379284259 hasFunder F4320323909 @default.
- W4379284259 hasLocation W43792842591 @default.
- W4379284259 hasOpenAccess W4379284259 @default.
- W4379284259 hasPrimaryLocation W43792842591 @default.
- W4379284259 hasRelatedWork W2041399278 @default.
- W4379284259 hasRelatedWork W2136184105 @default.
- W4379284259 hasRelatedWork W2160451891 @default.
- W4379284259 hasRelatedWork W3013515612 @default.
- W4379284259 hasRelatedWork W3186233728 @default.
- W4379284259 hasRelatedWork W4205958290 @default.
- W4379284259 hasRelatedWork W4327772909 @default.
- W4379284259 hasRelatedWork W4364301914 @default.
- W4379284259 hasRelatedWork W2187500075 @default.
- W4379284259 hasRelatedWork W2345184372 @default.
- W4379284259 hasVolume "231" @default.
- W4379284259 isParatext "false" @default.
- W4379284259 isRetracted "false" @default.
- W4379284259 workType "article" @default.