AI3SD Video: Outlier detection in Scientific Discovery
AI3SD Video: Outlier detection in Scientific Discovery
Detecting anomalous readings in data is a problem. Humans are good at some types, for example with images, however machines find it rather more difficult. Detecting anomalies in time series data is even more tricky. Discriminating between data that is part of the same distribution, or caused by some other process is also nontrivial. Anomaly detection is used in a wide range of applications, for example fraud detection for bank accounts, condition monitoring of mechanical systems, and in medical imagery. In all these applications, an outlier is indicative of a problem that requires further attention. A range of outlier detection methods is presented, and tested on a range of synthetic multivariate time series data. A novel method, cyclic regression, is presented and compared to more traditional methods. The application of these methods to real world data is demonstrated.
AI, AI3SD Event, Artificial Intelligence, Chemistry, COVID-19, Datasets, Machine Intelligence, Machine Learning, Materials Discovery, ML, Outliers, Research, Scientific Discovery
Grundy, Joanna
0bc72187-8dce-41fc-b809-93a6adbe0980
Kanza, Samantha
b73bcf34-3ff8-4691-bd09-aa657dcff420
Frey, Jeremy G.
ba60c559-c4af-44f1-87e6-ce69819bf23f
Niranjan, Mahesan
5cbaeea8-7288-4b55-a89c-c43d212ddd4f
Hooper, Victoria
af1a99f1-7848-4d5c-a4b5-615888838d84
20 January 2021
Grundy, Joanna
0bc72187-8dce-41fc-b809-93a6adbe0980
Kanza, Samantha
b73bcf34-3ff8-4691-bd09-aa657dcff420
Frey, Jeremy G.
ba60c559-c4af-44f1-87e6-ce69819bf23f
Niranjan, Mahesan
5cbaeea8-7288-4b55-a89c-c43d212ddd4f
Hooper, Victoria
af1a99f1-7848-4d5c-a4b5-615888838d84
Grundy, Joanna
(2021)
AI3SD Video: Outlier detection in Scientific Discovery.
Kanza, Samantha, Frey, Jeremy G., Niranjan, Mahesan and Hooper, Victoria
(eds.)
AI3SD Winter Seminar Series, , Online.
18 Nov 2020 - 21 Apr 2021 .
(doi:10.5258/SOTON/P0081).
Record type:
Conference or Workshop Item
(Other)
Abstract
Detecting anomalous readings in data is a problem. Humans are good at some types, for example with images, however machines find it rather more difficult. Detecting anomalies in time series data is even more tricky. Discriminating between data that is part of the same distribution, or caused by some other process is also nontrivial. Anomaly detection is used in a wide range of applications, for example fraud detection for bank accounts, condition monitoring of mechanical systems, and in medical imagery. In all these applications, an outlier is indicative of a problem that requires further attention. A range of outlier detection methods is presented, and tested on a range of synthetic multivariate time series data. A novel method, cyclic regression, is presented and compared to more traditional methods. The application of these methods to real world data is demonstrated.
Video
AI3SD-Winter-Seminar-Series-ML-JoGrundy-Final
- Version of Record
More information
Published date: 20 January 2021
Additional Information:
Graduated from Exeter University with a BSc in Chemistry in 1996, qualified as a teacher with a PGCE from Durham University then taught at Varndean School for a couple of years in Brighton, before doing an MSc in 1999, then DPhil in Organometallic chemistry at Sussex University. Between 2004 and 2006 I worked with Prof. Mathey at UC Riverside, before coming back to teach at a school in Duisburg Germany. I returned to the UK and taught chemistry and maths to A level, at Farnborough Sixth Form, then the Isle of Wight College, Barton Peveril and Itchen College. In 2018, did a second MSc in Artificial Intelligence at Southampton University, then worked as a Teaching Fellow in Computer Science for a year, before starting in my present job as a Research Fellow in Machine Learning.
Venue - Dates:
AI3SD Winter Seminar Series, , Online, 2020-11-18 - 2021-04-21
Keywords:
AI, AI3SD Event, Artificial Intelligence, Chemistry, COVID-19, Datasets, Machine Intelligence, Machine Learning, Materials Discovery, ML, Outliers, Research, Scientific Discovery
Identifiers
Local EPrints ID: 447701
URI: http://eprints.soton.ac.uk/id/eprint/447701
PURE UUID: e770b62d-e965-419b-9e9d-b91a6d94c093
Catalogue record
Date deposited: 18 Mar 2021 17:48
Last modified: 17 Mar 2024 03:55
Export record
Altmetrics
Contributors
Author:
Joanna Grundy
Editor:
Mahesan Niranjan
Editor:
Victoria Hooper
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics