Data Analytics Newsletter #35

Facebook whistleblower, Kaggle and hybrid work trends

The Data Analytics Practice Committee (DAPC), the Young Data Analytics Working Group (YDAWG) and the Actuaries Institute are pleased to bring you the latest in the world of Data Analytics to your inbox, and to share some of our recent work with you.

In this newsletter, we cover controversy at Facebook, precedents in patent law, hybrid work trends, and the Kaggle competition from earlier this year.

1. Ethics and regulation

**Facebook whistleblower Frances Haugen: The 60 Minutes interview**

Motivated by a friend lost to conspiracy theories, data scientist and former Facebook employee, Frances Haugen, has disclosed to media and regulators some concerning internal research findings. In this interview with 60 Minutes, Frances observes that content that evokes negative responses, such as anger or depression, also contributes the most engagement.

In doing so, Frances explains how Facebook is incentivised to optimise engagement for ad revenue and turn a blind eye to the consequences.

Another Facebook whistleblower has written a guide to whistleblowing. Read it here.

Deepmind health data scandal

Google’s DeepMind division is the target of an UK class-action lawsuit. It is alleged that DeepMind used confidential medical information from 1.6 million patients in a breach of data privacy. This follows the 2017 findings by the Information Commissioner that the Royal Free NHS Foundation Trust had improperly shared data with Deepmind.

The news comes as the UK looks to change its data laws away from the GDPR to “as light-touch a way as possible”.

AI can't own patents in the UK

The UK Court of Appeal has ruled that AI cannot be the inventor of a patent, as the associated intellectual property rights need to belong to a person. Disappointed? Don’t be - this follows more successful test cases in places like Australia.

2. Tools

Automate email reports with R and VBA

A step-by-step guide, with code, on how to automatically create reports and distribute them via email with R, VBA and scheduled tasks. Some good discussion of limitations highlighted in the comments (being that it runs on local machine).

Python decision tree visualisation

If you are a perfectionist about aesthetics, and have used scikit-learn for decision trees, you have probably been disappointed at the 'plot_tree' function. If so, consider dtreeviz as an alternative. See their links to Colab notebooks for examples.

3. Development

Kaggle winning solution analysis

The Actuaries Institute, Institute and Faculty of Actuaries and the Singapore Actuarial Society hosted a competition on Kaggle to promote development of data analytics talent, especially amongst actuaries. Curious about the successful approach? Read this Actuaries Digital article by Jin Cui, of the YDAWG, who helped make this competition happen.

4. Interesting reading

The bias-variance tradeoff

Why is it that machine learning models can have billions of parameters, and yet not overfit the data horribly? This paper provides a detailed, technical discussion on the topic, including an overview of the 'double-descent' phenomenon - where, as model size is increased, “performance first gets worse and then gets better”.

5. Strategic

Does AI play nice with humans?

An interesting article highlighting the human factor in situations where AI and humans need to work together. Researchers at the Massachusetts Institute of Technology found that whilst top AI algorithms of cooperative card game Hanabi play objectively well with other AI, human players struggled. One participant was “so stressed out at the ‘bad play’ from the top performing AI agent that they actually got a headache”.

Hybrid work infographics

The pandemic has led to the world’s largest working from home experiment, and it looks set to stay in some form. Many leaders now consider hybrid working to be the path forward. But some employees are struggling. Check out these infographics by Microsoft, highlighting some key trends.

Editors' note

Want to reclaim your privacy? Consider moving to a remote village.

As usual, check out Actuaries Digital for more things to read and the microsite for great learning resources and past Newsletter.

Jacky Poon, Henry Ma and Grant Lian
Editors, Data Analytics Newsletter

Disclaimer: The Institute wishes it to be understood that any opinions put forward in this publication are not necessarily those of the Institute.

Actuaries Institute
Level 2, 50 Carrington Street
Sydney NSW 2000, Australia
t +61 (0) 2 9239 6100

This email may contain privileged and/or confidential information. If you are not the intended recipient, please delete the email and notify the Actuaries Institute immediately on +61 (0) 2 9239 6100 or by return email. You must not disseminate, copy or take any action in reliance on the email. Neither any privilege nor confidentiality in the contents of this email is waived, lost or destroyed by reason that it has been transmitted other than to the intended addressee. If you send an email to us (including any emails addressed to a staff email address) the information in your email (including any ‘Personal Information’ as defined in the Privacy Act 1988 (Cth)) may be retained on our systems in accordance with our Privacy Policy and applicable data retention procedures.

Contents