UK ONS-UNECE Machine Learning Group 2022 webinar
November 30, 2022
New ML Group 2022 output page is now available through this link
Machine learning (ML) is drawing increasing attention among statistical organisations. ML can help carry out tasks that used to be carried out by humans such as reading and understanding textual data. This allows statistical organisations to perform manual tasks in a more efficient manner. ML is also an indispensable tool for processing and analysing the vast amount of information in big data, enabling the production of statistics on a more disaggregated level as well as new types of statistical products.
As statistical organisations explore ML they can benefit from sharing knowledge and experience, and from collaborating on developing common solutions. The ONS-UNECE Machine Learning Group 2022 is an initiative which brings together 420 members from 65 different organisations. The group has seven theme groups focusing on different ML application areas (web-scraping data, text classification, imagery data analysis, AIS data analysis) as well as cross-cutting issues arising from the integration of ML into the production processes of statistical organisations (quality of training data, IT infrastructure and model retraining).
Drawing on the group’s work in 2022, the webinar shared the results, lessons and discussions from the theme groups to offer insight into how machine learning can be applied to improve statistical production. It also explore some approaches to common challenges experienced in deploying ML models in statistical systems.
The webinar provided an opportunity for the statistical community to learn about the highlights of the group’s work in 2022 as well as the latest developments in machine learning in statistical organisations internationally.
The webinar recording is now available here .