Comment détecter des signaux faibles ? Un apport des data sciences à la lutte contre la fraude. (notice n° 1043841)
[ vue normale ]
| 000 -LEADER | |
|---|---|
| fixed length control field | 02864cam a2200289 4500500 |
| 005 - DATE AND TIME OF LATEST TRANSACTION | |
| control field | 20250125180325.0 |
| 041 ## - LANGUAGE CODE | |
| Language code of text/sound track or separate title | fre |
| 042 ## - AUTHENTICATION CODE | |
| Authentication code | dc |
| 100 10 - MAIN ENTRY--PERSONAL NAME | |
| Personal name | Laude, Henri |
| Relator term | author |
| 245 00 - TITLE STATEMENT | |
| Title | Comment détecter des signaux faibles ? Un apport des data sciences à la lutte contre la fraude. |
| 260 ## - PUBLICATION, DISTRIBUTION, ETC. | |
| Date of publication, distribution, etc. | 2016.<br/> |
| 500 ## - GENERAL NOTE | |
| General note | 36 |
| 520 ## - SUMMARY, ETC. | |
| Summary, etc. | Cet article propose une nouvelle méthode de détection de points aberrants dans une série temporelle multivariée qui découle directement de pratiques utilisées dans le cadre de ce que l’on nomme communément les data sciences. Les praticiens de l’intelligence économique traquent les signaux faibles dans des ensembles de données de plus en plus volumineux dont de nombreux attributs varient dans le temps. Les outils statistiques et graphiques couramment mis à leur disposition concernant les processus temporels s’appuient sur l’étude individuelle de chaque série et mettent en exergue les tendances et les variations par rapport à ces tendances. A l’inverse, notre méthode a pour objet d’aider l’analyste dans l’identification d’un signal lié à un comportement anormal eu égard à des relations, même difficilement détectables, entre de nombreuses séries temporelles.Une des applications naturelles de cette méthode est l’identification de fraudes ou de manipulations, en identifiant des signaux faibles dans les pistes d’audit de processus. |
| 520 ## - SUMMARY, ETC. | |
| Summary, etc. | This paper proposes a new method for outliers detection in multivariate time series which follows directly from practices used in the context of what is commonly called the data science. Practitioners need to chase the weak signals included in more and more voluminous data sets with many attributes changing over time. The statistical and graphical tools commonly available that are related to the temporal processes are based on individual study of each series and are dedicated to highlight trends and deviations from those trends. In contrast, our approach is to assist the analyst in identifying a signal related to aberrant behavior with regard to relationships, even difficult to detect, among many time series.One of the obvious applications of this method is the identification of fraud or psychological manipulations, identifying weak signals in audit trails process. |
| 690 ## - LOCAL SUBJECT ADDED ENTRY--TOPICAL TERM (OCLC, RLIN) | |
| Topical term or geographic name as entry element | data science |
| 690 ## - LOCAL SUBJECT ADDED ENTRY--TOPICAL TERM (OCLC, RLIN) | |
| Topical term or geographic name as entry element | points aberrants |
| 690 ## - LOCAL SUBJECT ADDED ENTRY--TOPICAL TERM (OCLC, RLIN) | |
| Topical term or geographic name as entry element | séries temporelles |
| 690 ## - LOCAL SUBJECT ADDED ENTRY--TOPICAL TERM (OCLC, RLIN) | |
| Topical term or geographic name as entry element | fraude |
| 690 ## - LOCAL SUBJECT ADDED ENTRY--TOPICAL TERM (OCLC, RLIN) | |
| Topical term or geographic name as entry element | détection de nouveautés |
| 690 ## - LOCAL SUBJECT ADDED ENTRY--TOPICAL TERM (OCLC, RLIN) | |
| Topical term or geographic name as entry element | outliers |
| 690 ## - LOCAL SUBJECT ADDED ENTRY--TOPICAL TERM (OCLC, RLIN) | |
| Topical term or geographic name as entry element | fraud |
| 690 ## - LOCAL SUBJECT ADDED ENTRY--TOPICAL TERM (OCLC, RLIN) | |
| Topical term or geographic name as entry element | novelty detection |
| 690 ## - LOCAL SUBJECT ADDED ENTRY--TOPICAL TERM (OCLC, RLIN) | |
| Topical term or geographic name as entry element | data science |
| 690 ## - LOCAL SUBJECT ADDED ENTRY--TOPICAL TERM (OCLC, RLIN) | |
| Topical term or geographic name as entry element | times series |
| 786 0# - DATA SOURCE ENTRY | |
| Note | Revue internationale d'intelligence économique | 8 | 1 | 2016-01-22 | p. 61-78 | 2101-647X |
| 856 41 - ELECTRONIC LOCATION AND ACCESS | |
| Uniform Resource Identifier | <a href="https://shs.cairn.info/revue-internationale-d-intelligence-economique-2016-1-page-61?lang=fr&redirect-ssocas=7080">https://shs.cairn.info/revue-internationale-d-intelligence-economique-2016-1-page-61?lang=fr&redirect-ssocas=7080</a> |
Pas d'exemplaire disponible.
Réseaux sociaux