Matches in SemOpenAlex for { <https://semopenalex.org/work/W1595561474> ?p ?o ?g. }
Showing items 1 to 70 of
70
with 100 items per page.
- W1595561474 abstract "Classifying detected flaws is becoming more and more important when applying NDT to modern complex structures and manufactured parts. Once a reflector within some mechanical component is detected, it is very important to know whether it is a sharp crack likely to lead to catastrophic failure. Classification in itself is nontrivial and it becomes a very difficult issue when large data volumes are acquired and large variations in measurement conditions are encountered. Thus it is highly desirable to have at least some decision support. Since there is human expertise available, it may seem natural to use Knowledge Based Systems. However, the problem of formulating the experience and knowledge of a human expert in general as explicit knowledge, is far from trivial in many cases (Waterman 1986), and experts in classification of ultrasonic signals is no exception (McNab 1995). It seems that neural networks, creating their own decision rules from examples, may be better suited for solving ultrasonic flaw classification problems. The problem of acquiring training examples of sufficient quality in sufficient quantities should be solvable by using numerical methods to predict the ultrasonic response from certain types of measurement situations in combination with measurements using real and induced flaws. For this report, a numerical model developed by Bostrom et al at Chalmers (Bostrom 1995) is the main source of data. The data volumes required to describe the response from one single flaw can be significant, e.g., using one full A-scan as input to a classifier may lead to a decision space with 1024 degrees of freedom or dimensions. The conventional approach to manage this is to simplify the classifier task by feature extraction. A literature study conducted previously has lead to the use of certain features (Eriksson 1994)." @default.
- W1595561474 created "2016-06-24" @default.
- W1595561474 creator A5014496992 @default.
- W1595561474 creator A5060635437 @default.
- W1595561474 date "1996-01-01" @default.
- W1595561474 modified "2023-09-27" @default.
- W1595561474 title "Characterization of Ultrasonic Signals Using Synthetic Data and Neural Networks" @default.
- W1595561474 cites W1518186130 @default.
- W1595561474 cites W2035123094 @default.
- W1595561474 doi "https://doi.org/10.1007/978-1-4419-8772-3_124" @default.
- W1595561474 hasPublicationYear "1996" @default.
- W1595561474 type Work @default.
- W1595561474 sameAs 1595561474 @default.
- W1595561474 citedByCount "0" @default.
- W1595561474 crossrefType "book-chapter" @default.
- W1595561474 hasAuthorship W1595561474A5014496992 @default.
- W1595561474 hasAuthorship W1595561474A5060635437 @default.
- W1595561474 hasConcept C119857082 @default.
- W1595561474 hasConcept C121332964 @default.
- W1595561474 hasConcept C124101348 @default.
- W1595561474 hasConcept C126838900 @default.
- W1595561474 hasConcept C153180895 @default.
- W1595561474 hasConcept C154945302 @default.
- W1595561474 hasConcept C24890656 @default.
- W1595561474 hasConcept C41008148 @default.
- W1595561474 hasConcept C50644808 @default.
- W1595561474 hasConcept C56529433 @default.
- W1595561474 hasConcept C71924100 @default.
- W1595561474 hasConcept C81288441 @default.
- W1595561474 hasConcept C95623464 @default.
- W1595561474 hasConceptScore W1595561474C119857082 @default.
- W1595561474 hasConceptScore W1595561474C121332964 @default.
- W1595561474 hasConceptScore W1595561474C124101348 @default.
- W1595561474 hasConceptScore W1595561474C126838900 @default.
- W1595561474 hasConceptScore W1595561474C153180895 @default.
- W1595561474 hasConceptScore W1595561474C154945302 @default.
- W1595561474 hasConceptScore W1595561474C24890656 @default.
- W1595561474 hasConceptScore W1595561474C41008148 @default.
- W1595561474 hasConceptScore W1595561474C50644808 @default.
- W1595561474 hasConceptScore W1595561474C56529433 @default.
- W1595561474 hasConceptScore W1595561474C71924100 @default.
- W1595561474 hasConceptScore W1595561474C81288441 @default.
- W1595561474 hasConceptScore W1595561474C95623464 @default.
- W1595561474 hasLocation W15955614741 @default.
- W1595561474 hasOpenAccess W1595561474 @default.
- W1595561474 hasPrimaryLocation W15955614741 @default.
- W1595561474 hasRelatedWork W131565041 @default.
- W1595561474 hasRelatedWork W1541633993 @default.
- W1595561474 hasRelatedWork W1868275169 @default.
- W1595561474 hasRelatedWork W2070521725 @default.
- W1595561474 hasRelatedWork W2077064445 @default.
- W1595561474 hasRelatedWork W2106116042 @default.
- W1595561474 hasRelatedWork W2157019209 @default.
- W1595561474 hasRelatedWork W2463059088 @default.
- W1595561474 hasRelatedWork W2555249502 @default.
- W1595561474 hasRelatedWork W2611097934 @default.
- W1595561474 hasRelatedWork W2761173574 @default.
- W1595561474 hasRelatedWork W2772025396 @default.
- W1595561474 hasRelatedWork W3008176958 @default.
- W1595561474 hasRelatedWork W50658105 @default.
- W1595561474 hasRelatedWork W69261196 @default.
- W1595561474 hasRelatedWork W86120781 @default.
- W1595561474 hasRelatedWork W95614390 @default.
- W1595561474 hasRelatedWork W985970753 @default.
- W1595561474 hasRelatedWork W2910175821 @default.
- W1595561474 hasRelatedWork W99132747 @default.
- W1595561474 isParatext "false" @default.
- W1595561474 isRetracted "false" @default.
- W1595561474 magId "1595561474" @default.
- W1595561474 workType "book-chapter" @default.