- February 20, 2019
1:00 pm - 4:00 pm
This Data Lab will be all about Anomaly detection (AD). Anomaly detection is the discovery of unexpected or interesting events, carried out by analyzing outliers or atypical data points in a dataset.
Anomaly detection as an activity piggybacks on a wide variety of underlying machine learning techniques – e.g. classification, clustering and distribution models. This means that there are a vast number of options for carrying out anomaly detection. As a result it is an area of machine learning that is both fascinating and challenging.
In a GOC context a SMALL number of applications could include (but this list is almost endless!)
- Fraud detection
- Financial & operations auditing
- Auditing data for data quality improvement
- Highlighting in real time critical issues
- Making operations more efficient by improving / automating decision making
- Data security
- Forecasting events
- Image processing
- Quality Assurance (e.g. in text data)
- Damage detection
- etc, etc, etc
We encourage you to think about how AD might be helpful in your specific job role / context and bring along some examples that we can explore together as a group.
The agenda for the day will be:
- 1:00 pm – meet, greet and chat
- 1:15 pm – welcomes and updates
- 1:20 pm – AI in the news & discussion
- 1:45 pm – 4:00 pm (with a break in the middle) – presentation on AD and a general discussion on application to specific examples from the attendees
For a first time attendee the event is free (up to 3 people from new organization). For other participants a subscription is required BUT as we are still setting up our registration process please just register for now and we will sort that out later on.
Venue: The AIM Group