The HLG-MOS Machine Learning Project webinar will be held virtually
from 16 to 17 November (13:30-17:30 CET, Geneva time)
Project outputs are available on this wiki page
The webinar will be held via Webex. Please follow instructions in How to Join WebEx to join the webinar.
- When you join WebEx, you will be asked to provide "name" and "email", please use country/organisation, first name and last name as "name" (e.g. UNECE_InKyung Choi). If you want to change the name after log-in, please check this instruction
- For smooth proceeding, please mute microphone (except when you speak) and turn off camera.
- If you don’t hear any sound, the sound quality is bad, or others can't hear you, please double check your device connection. See detailed instructions at Webex Audio Troubleshooting.
- If you have any problem or question about connection during the webinar, please send a chat message to UNECE
* time is in Central European Time (CET), Geneva time
** programme is subject to change, all documents will be uploaded as as they become available
Day 1 - November 16
Time (CET) | Presentation | Speakers | Documents |
13:30 - 13:40 | Welcome and introduction to ML project & Pilot studies | Stéphane Dufour (Statistics Canada and Co-chair of HLG-MOS Executive Board) Claude Julien (UNECE) Eric Deeben (UK, ONS Data Science Campus) | |
Work Package 1. Pilot Study: Theme - Coding & Classification | |||
| 13:40 - 13:55 | Occupation and Economic activity coding using natural language processing | Jael Pérez Sànchez (INEGI, Mexico) | |
| 13:55 - 14:10 | Demonstration with ML code and data | Krystyna Piatkowska & Marta Kruczek-Szepel (Statistics Poland) | |
14:10 - 14:20 | Questions and answers | ||
| 14:20 - 14:40 | Coding and Classification theme report | Claus Sthamer (UK, ONS Data Science Campus) | |
| 14:40 - 15:00 | Questions, answers and discussion | ||
15:00 - 15:10 | Break | ||
Work Package 1. Pilot Study: Theme - Editing and Imputation | |||
15:10 - 15:25 | Imputation of the variable “Attained Level of Education” in Base Register of Individuals | Fabrizio De Fausti (Istat, Italy) | |
| 15:25 - 15:40 | Machine Learning for Data Editing Cleaning in NSI : Some ideas and hints | Fabiana Rocci (Istat, Italy) | |
| 15:40 - 15:50 | Questions and answers | ||
| 15:50 - 16:10 | Editing and Imputation theme report | Florian Dumpert (Destatis, Germany) | |
| 16:10 - 16:30 | Questions, answers and discussion | ||
16:30 - 16:40 | Break | ||
Work Package 2. Quality | |||
16:40 - 17:10 | Quality Framework for Statistical Algorithms (QF4SA) | Wesley Yung (Statistics Canada) | |
| Explainability | Joep Berger (Statistics Netherlands) | ||
| 17:10 - 17:30 | Questions, answers and discussion | ||
Day 2 - November 17
Time (CET) | Presentation | Speaker | Documents |
|---|---|---|---|
Work Package 1. Pilot Study: Imagery | |||
| 13:30 - 13:45 | Generic Pipeline for Production of Official Statistics Using Satellite Data and Machine Learning | InKyung Choi (UNECE) | |
| 13:45 - 14:00 | Address Register Automated Image Recognition (AIR) model | Daniel Merkas & James Farnell (Australian Bureau of Statistics) | |
| 14:00 - 14:10 | Questions and answers | ||
| 14:10 - 14:30 | Imagery theme report | Abel Coronado and Jimena Juàrez Carrillo (INEGI, Mexico) | |
| 14:30 - 14:50 | Questions, answers and discussion | ||
14:50 - 15:05 | Pilot studies - summary | Eric Deeben (ONS Data Science Campus, UK) | |
15:05 - 15:15 | Break | ||
Work Package 3. Integration of Machine Learning | |||
| 15:15 - 15:40 | Integration of Machine Learning | Alex Measure (Bureau of Labor Statistics, USA) | |
| 15:40 - 16:00 | Questions, answers and discussion | ||
| Project output, conclusion and future | |||
16:00 - 16:15 | Machine Learning Project Outputs | InKyung Choi & Claude Julien (UNECE) | |
16:15 - 16:30 | Conclusion | Claude Julien (UNECE) | |
16:30 - 16:45 | Break | ||
16:45 - 17:30 | Future directions | Eric Deeben (ONS Data Science Campus, UK) | |
With rapidly growing interest in the use of machine learning for official statistics but with limited experience with concrete applications, there was a great need for a common platform where experts in national statistics offices to test their ideas, exchange experiences and collaborate on developments. National statistics offices work on similar type of problems and operate with similar business constraints, hence can benefit from developing shared understanding. To address this need, the High-Level Group on Modernisation of Official Statistics (HLG-MOS) launched the Machine Learing project in early 2019 with aims to:
- Investigate and demonstrate the value added of ML in the production of official statistics, where "value added" is increase in relevance, better overall quality or reduction in costs;
- Advance the capability of ML to add value to the production of official statistics;
- Advance the capability of national statistical organisations to use ML in the production of official statistics;
- Enhance collaboration between statistical organisations in the development and application of ML.
Following these objectives, the project team identified three main areas to advance the use of ML in statistical organisations:
- Work package 1 – Pilot Studies (demonstration of value added)
- Coding and Classification
- Edit and Imputation
- Imagery
- Work package 2 – Quality
- Work package 3 – Integration of ML into organisation
This webinar is the first public event where the outputs of the project will be communicated. This includes study reports, shared code and data, analysis of value added, recommended ML practices, quality framework elements and examples of organisational practices to address integration challenges.
The project will officially close at the end of the year (2020). Since it was launched in March 2019, the number of participants and other collaborators has grown from 20 to over 120. Given this strong interest, the project will evolve into a group to continue the advancement of ML in the production of official statistics.