TITLE: EXPLAIN 4.0: DEVELOPMENT OF INTERPRETABLE ARTIFICIAL INTELLIGENCE SOLUTIONS FOR INDUSTRY 4.0 (Q4242250)
Jump to navigation
Jump to search
Project Q4242250 in Italy
| Language | Label | Description | Also known as |
|---|---|---|---|
| English | TITLE: EXPLAIN 4.0: DEVELOPMENT OF INTERPRETABLE ARTIFICIAL INTELLIGENCE SOLUTIONS FOR INDUSTRY 4.0 |
Project Q4242250 in Italy |
Statements
17,415.0 Euro
0 references
34,830.0 Euro
0 references
50.0 percent
0 references
STATWOLF DATA SCIENCE S.R.L.
0 references
EXPLAIN 4.0 MIRA A SVILUPPARE SISTEMI DI SUPPORTO ALLE DECISIONI (SSD) BASATE SU TECNICHE DI MACHINE LEARNING (ML) INTERPRETABILI PER L - INDUSTRIA 4.0. UNO DEI MAGGIORI SCOGLI PER LA DIFFUSIONE DI SOLUZIONI BASATE SU ML IN AMBITO INDUSTRIALE (E NON) E' INFATTI LA SCARSA FIDUCIA NEI CONFRONTI DELLE PREDIZIONI DEGLI ALGORITMI, SPECIALMENTE DA PARTE DI PERSONE NON FORMATE SU TALI TEMATICHE. IL PROGETTO, GRAZIE AL RECLUTAMENTO DI UN GIOVANE RICERCATORE CON UN PROFILO DA DATA SCIENTIST, SI CONCENTRERA' SU TECNICHE DI ML NON SUPERVISIONATE; QUESTE HANNO UN - ENORME RILEVANZA IN AMBITO INDUSTRIALE ABILITANDO SOLUZIONI PER IL RICONOSCIMENTO DI ANOMALIE CHE HANNO IL VANTAGGIO DI NON ESSERE BASATE SU `ETICHETTATURE - DEL DATO DA PARTE DI UN ESSERE UMANO, CHE SONO SPESSO NON PRESENTI O POCO AFFIDABILI IN AMBITI REALI. NEL CONTESTO DEL RICONOSCIMENTO DELLE ANOMALIE, L - INTERPRETABILITA' DI TALI SOLUZIONI, OLTRE AD INCREMENTARE L - ADOZIONE DA PARTE DELL - UTENTE, HA ANCHE L - ENORME VANTAGGIO D (Italian)
0 references
EXPLAIN 4.0 AIMS TO DEVELOP DECISION SUPPORT SYSTEMS (SSDS) BASED ON MACHINE LEARNING TECHNIQUES (ML) INTERPRETABLE FOR INDUSTRY 4.0. ONE OF THE BIGGEST REEFS FOR THE DISSEMINATION OF ML-BASED SOLUTIONS IN INDUSTRY (AND NOT) IS THE LACK OF CONFIDENCE IN ALGORITHM PREDICTIONS, ESPECIALLY FROM PEOPLE NOT TRAINED ON THESE ISSUES. THE PROJECT, THANKS TO THE RECRUITMENT OF A YOUNG RESEARCHER WITH A SCIENTIST DATA PROFILE, WILL FOCUS ON UNSUPERVISED ML TECHNIQUES; THESE HAVE A HUGE INDUSTRIAL RELEVANCE BY ENABLING SOLUTIONS FOR THE RECOGNITION OF ANOMALIES THAT HAVE THE ADVANTAGE OF NOT BEING BASED ON ‘LABELLING’ OF DATA BY A HUMAN BEING, WHICH ARE OFTEN NOT PRESENT OR UNRELIABLE IN REAL AREAS. IN THE CONTEXT OF THE RECOGNITION OF ANOMALIES, THE INTERPRETABILITY OF THESE SOLUTIONS, IN ADDITION TO INCREASING THE — ADOPTION BY THE USER, ALSO HAS THE ENORMOUS ADVANTAGE OF (English)
2 February 2022
0 references
L’EXPLICATION 4.0 VISE À DÉVELOPPER DES SYSTÈMES D’AIDE À LA DÉCISION (SSD) BASÉS SUR DES TECHNIQUES D’APPRENTISSAGE AUTOMATIQUE (ML) INTERPRÉTABLES POUR L’INDUSTRIE 4.0. L’UN DES PLUS GRANDS RÉCIFS POUR LA DIFFUSION DE SOLUTIONS À BASE DE ML DANS L’INDUSTRIE (ET NON) EST LE MANQUE DE CONFIANCE DANS LES PRÉDICTIONS ALGORITHMIQUES, EN PARTICULIER DE LA PART DE PERSONNES QUI N’ONT PAS REÇU DE FORMATION SUR CES QUESTIONS. LE PROJET, GRÂCE AU RECRUTEMENT D’UN JEUNE CHERCHEUR AYANT UN PROFIL DE DONNÉES SCIENTIFIQUES, SE CONCENTRERA SUR LES TECHNIQUES ML NON SURVEILLÉES; CELLES-CI ONT UNE IMPORTANCE INDUSTRIELLE CONSIDÉRABLE EN PERMETTANT DES SOLUTIONS POUR LA RECONNAISSANCE D’ANOMALIES QUI ONT L’AVANTAGE DE NE PAS ÊTRE BASÉES SUR L’«ÉTIQUETAGE» DES DONNÉES PAR UN ÊTRE HUMAIN, QUI SONT SOUVENT INEXISTANTES OU PEU FIABLES DANS DES DOMAINES RÉELS. DANS LE CONTEXTE DE LA RECONNAISSANCE DES ANOMALIES, L’INTERPRÉTABILITÉ DE CES SOLUTIONS, EN PLUS D’AUGMENTER L’ADOPTION PAR L’UTILISATEUR, A ÉGALEMENT L’ÉNORME AVANTAGE DE: (French)
2 February 2022
0 references
PADOVA
0 references