Skip to main content

Workshop on Modernization of Official Statistics, online

Workshop on Modernization of Official Statistics, online

15 - 16 November 2021

The purpose of this workshop is to ensure that the work is community driven and that activities and initiatives are aligned with the implementation of the HLG‑MOS vision, avoiding duplication and maximizing efficiency. The expected outputs are a set of agreed and prioritized implementation actions.

The target audience are all participants in the 2021 HLG-MOS work programme and statisticians with technical knowledge of modernisation of statistics combined with a broader understanding of the statistical process and its modernisation, including some knowledge about international developments in this area.

This workshop was extended with Webinars of the two 2021 HLG-MOS projects on Synthetic Data Sets and on Input Privacy-Preservation Techniques.

Monday, 15 November 2021

  Project
Proposal
Presentations
Opening and Welcome: Lidia Bratanova and Taeke Gjaltema (UNECE), Stéphane Dufour (Statistics Canada, co-chair HLG-MOS Executive Board) and Jennifer Banim (CSO Ireland, co-chair HLG-MOS Executive Board)
Keynote Speech: Anil Arora chair of the HLG-MOS   PDF
Input Privacy-Preservation Techniques Project. Dennis Ramondt (Statistics Netherlands and UNECE Project Manager)   PDF
Synthetic Data Project. Kate Burnett-Isaacs (Statistics Canada and UNECE Project Manager)   PDF

Proposal 1: Project extension for Input Privacy Preserving Techniques

Dennis Ramondt (Statistics Netherlands and UNECE Project Manager)

PDF

PDF

Proposal 2: Project proposal for Data Governance Framework to Achieve Data Interoperability

Juan Muñoz López (INEGI, Mexico)

PDF

PDF

Proposal 3:  Project proposal for Meta Academy

Kate Burnett-Isaacs (Statistics Canada) and Eric Anvar (OECD)

PDF

PDF

Tuesday, 16 November 2021

Opening and Welcome: Taeke Gjaltema (UNECE), Stéphane Dufour (Statistics Canada, co-chair HLG-MOS Executive Board) and Jennifer Banim (CSO Ireland, co-chair HLG-MOS Executive Board)

Outcome Expert Meetings/Workshops. Taeke Gjaltema (UNECE)

Presentation

Machine Learning Community 2021. Eric Deeben (ONS)

Presentation

Supporting Standards. Zoltán Vereczkei, chair (Statistics Hungary)

Outputs:

Presentation

Activity proposals:

Capabilities and Communication. Anna Borowska (Statistics Poland) and Maria Hurley (CSO Ireland) co-chairs

Outputs:

Presentation

Activity proposals:

Blue Skies Thinking Network: HLG-MOS Proposals through BSTN for 2022. Barteld Braaksma, chair (Statistics Netherlands)

Presentation

Document

Other Activity Proposals and Soapbox/Pitch talks

  • Join the VTL community
  • Data Virtualization community

 

New Modernisation Group: Applying Data Science and Modern Methods. Stéphane Dufour (Statistics Canada, co-chair HLG-MOS Executive Board)

 
Discussion New Group: areas of work  
Conclusions and way forward. Stéphane Dufour (Statistics Canada, co-chair HLG-MOS Executive Board) and Jennifer Banim (CSO Ireland, co-chair HLG-MOS Executive Board)  

Closing Workshop

 

Synthetic Data Project Webinar - 17 November 2021

Introduction Kate Burnett-Isaacs (Statistics Canada and UNECE Project Manager)  

Chapter 2: Use Cases Kate Burnett-Isaacs (Statistics Canada and UNECE Project Manager)

presentation

Chapter 3: Methods and recommendations Kenza Sallier (Statistics Canada)

presentation

Break  

Chapter 5: Utility Measures and recommendations Kate Burnett-Isaacs (Statistics Canada and UNECE Project Manager), Gillian Raab (Emeritus Professor, Edinburgh Napier University, Part-time Research Fellow Administrative Data Research Centre - Scotland) and Christine Task (Knexus Research)

presentation

Wrap up and next steps Kate Burnett-Isaacs (Statistics Canada and UNECE Project Manager)

 

Closing Webinar

 

Input Privacy-Preservation Project Webinar - 18 November 2021

Introduction to the project Dennis Ramondt presentation

Input Privacy: Towards a Logical Framework for Defining Official Statistics Scenarios

Monica Scannapieco presentation

Privacy Set Intersection with Analytics

Massimo De Cubellis

presentation
Private Set Intersection Use Case and Synthetic Data Abel Dasylva presentation

Private set intersection with analytics and use of Homomorphic Encryption

Ralph Schreijen presentation
Private machine learning Saeid Molladavoudi presentation

Secure Private Computing-as-a-service

Fabio Ricciato presentation
Input Privacy preserving techniques Project Outputs (next step) Dennis Ramondt presentation

Machine Learning Group 2021 Webinar - 19 November 2021

Opening and Setting the Scene

Welcome and Project Overview

- Eric Deeben (Head of International, Data Science Campus, ONS, UK)

- Louisa Nolan (Chief Data Scientist, Data Science Campus, ONS, UK)

 

Guest Panel: Strategic Priorities for Machine Learning in the work of National Statistics Offices

- Louisa Nolan (ONS, UK)

- Sevgui Erman (Statistics Canada; Director and Chief Data Scientist)

- Jakob Engdahl (Statistics Sweden; Senior Strategist)

 

Work Stream 1. Pilot studies: from Idea to Valid Solutions

The Utilisation of Satellite Imagery Analysis for Poverty Mapping in Indonesia 

- Arie Wahyu (Statistics Indonesia)

Presentation

Exploring Model-Assisted Machine Learning Estimates of Consumer Expenditures

- Clayton Knappenberger (US, Census Bureau) 

Presentation

A Shared Service for Text Classification

- Klaus Lehmann (Chile, INE)

Presentation

Work Stream 2. From Valid Solution to Production / Work Stream 3. Data Ethics and Governance

Use Cases applied to a Data Lake Prototype in a National Statistical Office 

- Abel Coronado (Mexico, INEGI)

Presentation

Ethical Consideration in the Use of Machine Learning for Research and Statistics 

- Alice Toms (UK, Statistics Authority)

Presentation

Work Stream 4. On the Quality of Training Data / Work Stream 5. On the Quality Framework for Statistical Algorithm 

From Theory to Practice: Detecting Model Decay

- Riitta Piela (Statistics Finland)

Presentation

Holistic Approaches to Evaluating a Machine Learning Model: A case study for automated coding 

- Jose Jiminez (Mexico, INEGI)

Presentation

Machine Learning Group - the way forward

Future work presentation

- Alison Baily (UK, ONS) and InKyung Choi (UNECE)

 

23237 _ ML2021_Webinar_Indonesia _ 364245 _ English _ 773 _ 346072 _ pdf

Opening and Setting the Scene

Welcome and Project Overview

- Eric Deeben (Head of International, Data Science Campus, ONS, UK)

- Louisa Nolan (Chief Data Scientist, Data Science Campus, ONS, UK)

 

Guest Panel: Strategic Priorities for Machine Learning in the work of National Statistics Offices

- Louisa Nolan (ONS, UK)

- Sevgui Erman (Statistics Canada; Director and Chief Data Scientist)

- Jakob Engdahl (Statistics Sweden; Senior Strategist)

 

Work Stream 1. Pilot studies: from Idea to Valid Solutions

The Utilisation of Satellite Imagery Analysis for Poverty Mapping in Indonesia 

- Arie Wahyu (Statistics Indonesia)

Presentation

Exploring Model-Assisted Machine Learning Estimates of Consumer Expenditures

- Clayton Knappenberger (US, Census Bureau) 

Presentation

A Shared Service for Text Classification

- Klaus Lehmann (Chile, INE)

Presentation

Work Stream 2. From Valid Solution to Production / Work Stream 3. Data Ethics and Governance

Use Cases applied to a Data Lake Prototype in a National Statistical Office 

- Abel Coronado (Mexico, INEGI)

Presentation

Ethical Consideration in the Use of Machine Learning for Research and Statistics 

- Alice Toms (UK, Statistics Authority)

Presentation

Work Stream 4. On the Quality of Training Data / Work Stream 5. On the Quality Framework for Statistical Algorithm 

From Theory to Practice: Detecting Model Decay

- Riitta Piela (Statistics Finland)

Presentation

Holistic Approaches to Evaluating a Machine Learning Model: A case study for automated coding 

- Jose Jiminez (Mexico, INEGI)

Presentation

Machine Learning Group - the way forward

Future work presentation

- Alison Baily (UK, ONS) and InKyung Choi (UNECE)

 

23237 _ ML2021_Webinar_US _ 364246 _ English _ 773 _ 346079 _ pdf

Opening and Setting the Scene

Welcome and Project Overview

- Eric Deeben (Head of International, Data Science Campus, ONS, UK)

- Louisa Nolan (Chief Data Scientist, Data Science Campus, ONS, UK)

 

Guest Panel: Strategic Priorities for Machine Learning in the work of National Statistics Offices

- Louisa Nolan (ONS, UK)

- Sevgui Erman (Statistics Canada; Director and Chief Data Scientist)

- Jakob Engdahl (Statistics Sweden; Senior Strategist)

 

Work Stream 1. Pilot studies: from Idea to Valid Solutions

The Utilisation of Satellite Imagery Analysis for Poverty Mapping in Indonesia 

- Arie Wahyu (Statistics Indonesia)

Presentation

Exploring Model-Assisted Machine Learning Estimates of Consumer Expenditures

- Clayton Knappenberger (US, Census Bureau) 

Presentation

A Shared Service for Text Classification

- Klaus Lehmann (Chile, INE)

Presentation

Work Stream 2. From Valid Solution to Production / Work Stream 3. Data Ethics and Governance

Use Cases applied to a Data Lake Prototype in a National Statistical Office 

- Abel Coronado (Mexico, INEGI)

Presentation

Ethical Consideration in the Use of Machine Learning for Research and Statistics 

- Alice Toms (UK, Statistics Authority)

Presentation

Work Stream 4. On the Quality of Training Data / Work Stream 5. On the Quality Framework for Statistical Algorithm 

From Theory to Practice: Detecting Model Decay

- Riitta Piela (Statistics Finland)

Presentation

Holistic Approaches to Evaluating a Machine Learning Model: A case study for automated coding 

- Jose Jiminez (Mexico, INEGI)

Presentation

Machine Learning Group - the way forward

Future work presentation

- Alison Baily (UK, ONS) and InKyung Choi (UNECE)

 

23237 _ ML2021_Webinar_Chile _ 364247 _ English _ 773 _ 346080 _ pdf

Opening and Setting the Scene

Welcome and Project Overview

- Eric Deeben (Head of International, Data Science Campus, ONS, UK)

- Louisa Nolan (Chief Data Scientist, Data Science Campus, ONS, UK)

 

Guest Panel: Strategic Priorities for Machine Learning in the work of National Statistics Offices

- Louisa Nolan (ONS, UK)

- Sevgui Erman (Statistics Canada; Director and Chief Data Scientist)

- Jakob Engdahl (Statistics Sweden; Senior Strategist)

 

Work Stream 1. Pilot studies: from Idea to Valid Solutions

The Utilisation of Satellite Imagery Analysis for Poverty Mapping in Indonesia 

- Arie Wahyu (Statistics Indonesia)

Presentation

Exploring Model-Assisted Machine Learning Estimates of Consumer Expenditures

- Clayton Knappenberger (US, Census Bureau) 

Presentation

A Shared Service for Text Classification

- Klaus Lehmann (Chile, INE)

Presentation

Work Stream 2. From Valid Solution to Production / Work Stream 3. Data Ethics and Governance

Use Cases applied to a Data Lake Prototype in a National Statistical Office 

- Abel Coronado (Mexico, INEGI)

Presentation

Ethical Consideration in the Use of Machine Learning for Research and Statistics 

- Alice Toms (UK, Statistics Authority)

Presentation

Work Stream 4. On the Quality of Training Data / Work Stream 5. On the Quality Framework for Statistical Algorithm 

From Theory to Practice: Detecting Model Decay

- Riitta Piela (Statistics Finland)

Presentation

Holistic Approaches to Evaluating a Machine Learning Model: A case study for automated coding 

- Jose Jiminez (Mexico, INEGI)

Presentation

Machine Learning Group - the way forward

Future work presentation

- Alison Baily (UK, ONS) and InKyung Choi (UNECE)

 

23237 _ ML2021_Webinar_Mexico_DataLake _ 364248 _ English _ 773 _ 346081 _ pdf

Opening and Setting the Scene

Welcome and Project Overview

- Eric Deeben (Head of International, Data Science Campus, ONS, UK)

- Louisa Nolan (Chief Data Scientist, Data Science Campus, ONS, UK)

 

Guest Panel: Strategic Priorities for Machine Learning in the work of National Statistics Offices

- Louisa Nolan (ONS, UK)

- Sevgui Erman (Statistics Canada; Director and Chief Data Scientist)

- Jakob Engdahl (Statistics Sweden; Senior Strategist)

 

Work Stream 1. Pilot studies: from Idea to Valid Solutions

The Utilisation of Satellite Imagery Analysis for Poverty Mapping in Indonesia 

- Arie Wahyu (Statistics Indonesia)

Presentation

Exploring Model-Assisted Machine Learning Estimates of Consumer Expenditures

- Clayton Knappenberger (US, Census Bureau) 

Presentation

A Shared Service for Text Classification

- Klaus Lehmann (Chile, INE)

Presentation

Work Stream 2. From Valid Solution to Production / Work Stream 3. Data Ethics and Governance

Use Cases applied to a Data Lake Prototype in a National Statistical Office 

- Abel Coronado (Mexico, INEGI)

Presentation

Ethical Consideration in the Use of Machine Learning for Research and Statistics 

- Alice Toms (UK, Statistics Authority)

Presentation

Work Stream 4. On the Quality of Training Data / Work Stream 5. On the Quality Framework for Statistical Algorithm 

From Theory to Practice: Detecting Model Decay

- Riitta Piela (Statistics Finland)

Presentation

Holistic Approaches to Evaluating a Machine Learning Model: A case study for automated coding 

- Jose Jiminez (Mexico, INEGI)

Presentation

Machine Learning Group - the way forward

Future work presentation

- Alison Baily (UK, ONS) and InKyung Choi (UNECE)

 

23237 _ ML2021_Webinar_Finland _ 364249 _ English _ 773 _ 346082 _ pdf

Opening and Setting the Scene

Welcome and Project Overview

- Eric Deeben (Head of International, Data Science Campus, ONS, UK)

- Louisa Nolan (Chief Data Scientist, Data Science Campus, ONS, UK)

 

Guest Panel: Strategic Priorities for Machine Learning in the work of National Statistics Offices

- Louisa Nolan (ONS, UK)

- Sevgui Erman (Statistics Canada; Director and Chief Data Scientist)

- Jakob Engdahl (Statistics Sweden; Senior Strategist)

 

Work Stream 1. Pilot studies: from Idea to Valid Solutions

The Utilisation of Satellite Imagery Analysis for Poverty Mapping in Indonesia 

- Arie Wahyu (Statistics Indonesia)

Presentation

Exploring Model-Assisted Machine Learning Estimates of Consumer Expenditures

- Clayton Knappenberger (US, Census Bureau) 

Presentation

A Shared Service for Text Classification

- Klaus Lehmann (Chile, INE)

Presentation

Work Stream 2. From Valid Solution to Production / Work Stream 3. Data Ethics and Governance

Use Cases applied to a Data Lake Prototype in a National Statistical Office 

- Abel Coronado (Mexico, INEGI)

Presentation

Ethical Consideration in the Use of Machine Learning for Research and Statistics 

- Alice Toms (UK, Statistics Authority)

Presentation

Work Stream 4. On the Quality of Training Data / Work Stream 5. On the Quality Framework for Statistical Algorithm 

From Theory to Practice: Detecting Model Decay

- Riitta Piela (Statistics Finland)

Presentation

Holistic Approaches to Evaluating a Machine Learning Model: A case study for automated coding 

- Jose Jiminez (Mexico, INEGI)

Presentation

Machine Learning Group - the way forward

Future work presentation

- Alison Baily (UK, ONS) and InKyung Choi (UNECE)

 

23237 _ ML2021_Webinar_UK _ 364250 _ English _ 773 _ 346084 _ pdf

Opening and Setting the Scene

Welcome and Project Overview

- Eric Deeben (Head of International, Data Science Campus, ONS, UK)

- Louisa Nolan (Chief Data Scientist, Data Science Campus, ONS, UK)

 

Guest Panel: Strategic Priorities for Machine Learning in the work of National Statistics Offices

- Louisa Nolan (ONS, UK)

- Sevgui Erman (Statistics Canada; Director and Chief Data Scientist)

- Jakob Engdahl (Statistics Sweden; Senior Strategist)

 

Work Stream 1. Pilot studies: from Idea to Valid Solutions

The Utilisation of Satellite Imagery Analysis for Poverty Mapping in Indonesia 

- Arie Wahyu (Statistics Indonesia)

Presentation

Exploring Model-Assisted Machine Learning Estimates of Consumer Expenditures

- Clayton Knappenberger (US, Census Bureau) 

Presentation

A Shared Service for Text Classification

- Klaus Lehmann (Chile, INE)

Presentation

Work Stream 2. From Valid Solution to Production / Work Stream 3. Data Ethics and Governance

Use Cases applied to a Data Lake Prototype in a National Statistical Office 

- Abel Coronado (Mexico, INEGI)

Presentation

Ethical Consideration in the Use of Machine Learning for Research and Statistics 

- Alice Toms (UK, Statistics Authority)

Presentation

Work Stream 4. On the Quality of Training Data / Work Stream 5. On the Quality Framework for Statistical Algorithm 

From Theory to Practice: Detecting Model Decay

- Riitta Piela (Statistics Finland)

Presentation

Holistic Approaches to Evaluating a Machine Learning Model: A case study for automated coding 

- Jose Jiminez (Mexico, INEGI)

Presentation

Machine Learning Group - the way forward

Future work presentation

- Alison Baily (UK, ONS) and InKyung Choi (UNECE)

 

23237 _ ML2021_Webinar_Mexico_Quality _ 364251 _ English _ 773 _ 346086 _ pdf