Data Science Insider: March 19th, 2021

In this week’s Super Data Science newsletter: Human Rights Trust Raises Concerns About AI. AI Project to ‘Pandemic-Proof’ NHS Supply Chain. ML and Causality. Advanced Techniques for Coding Python. Instagram Turns to AI to Protect Children. Cheers, - The SuperDataScience Team P.S. Have friends and colleagues who could benefit from these weekly updates? Send them to [this link]( to subscribe to the Data Science Insider. --------------------------------------------------------------- [Human Rights Trust Raises Concerns About AI]( brief: Fears have been raised by human rights groups about the ways that AI and facial recognition software is being used by the government in Myanmar against their own citizens. This Thomson Reuters Foundation article highlights concerns that Chinese technology from Huawei is being used to track protestors. The Human Rights Trust say the use of AI to check on citizens' movements poses a "serious threat" to their liberty and that the country risks implementing a "digital dictatorship." Myanmar has been in turmoil since the military ousted civilian leader Aung San Suu Kyi on 1 February, triggering a mass uprising that security forces have sought to crush with a campaign of violence. Since then, hundreds of CCTV cameras have been installed, leading to "heightened concern" over cameras armed with AI technology that can scan faces and vehicle licence plates in public places, and alert authorities to those on a wanted list. Why this is important: Since the coup, more than 220 people have been confirmed killed and 2,000 detained, according to a local monitoring group. The use of facial recognition software is a worrying development in the military junta’s arsenal of weapons against protestors. [Click here to find out!]( [AI Project to ‘Pandemic-Proof’ NHS Supply Chain]( brief: A collaboration between Sheffield University and AI healthcare marketplace Vamstar is aiming to help the NHS to manage its supply chain more efficiently. The project team believes that the platform will help to prevent future shortages of essential products, such as the shortage of PPE reported by many health and social care organisations during the first wave of the Covid-19 pandemic. With the ability to analyse NHS and global procurement data from previous supply contracts, the platform will aim to allow NHS buyers to evaluate credibility and capability of suppliers to fulfil their order. Researchers at Sheffield University’s Information School are said to be developing Natural Language Processing (NLP) methods for the automated reading and extraction of data from large amounts of contract tender data held by the NHS and other European healthcare providers. Vamstar will work alongside them to incorporate the information into a healthcare procurement marketplace. Why this is important: As we (finally!) appear to be coming out of the other side of the coronavirus pandemic, it is now time to learn lessons from the past year. As an industry we can have a large role to play in ensuing that we never suffer the same way again. [Click here to read on!]( [ML and Causality]( In brief: This article is part of a series from BD Tech Talks which analyses and reviews AI research papers. This article discusses a paper titled “Towards Causal Representation Learning” in which researchers at the Max Planck Institute for Intelligent Systems, the Montreal Institute for Learning Algorithms (Mila), and Google Research, examines the challenges that arise from the lack of causal representations in ML models and provide directions for creating ML systems that can learn causal representations. The article argues that ML has a lack of causality, which is responsible many of the major challenges the field faces. Instead, it states that the ML community, as a whole, is too focused on solving problems of independent and identically distributed (IID) data and too little on learning causal representations, stating: “[ML] often disregards information that animals use heavily: interventions in the world, domain shifts, temporal structure […]- we consider these factors a nuisance.” Why this is important: The ML community’s focus on IID over causality has made headlines over recent years, in instances such as self-driving cars making dangerous decisions. Although the ideas in this paper are at the conceptual stage, building causality representations into ML may be something we will see more of in the future. [Click here to discover more!]( [Advanced Techniques for Coding Python]( In brief: In this article from Better Programming, Physicist, Machine Learning Scientist and Software Engineer, Dr Bruce H. Cottman uses his experience to give thirty-five tips for better documentation, coding, testing, verification, and continuous integration in Python. Cottman divides the techniques that he’s learned over his lengthy career into five categories: Documentation techniques, coding techniques, testing techniques, verification techniques and continuous integration (CI) techniques. Cottman uses PHOTONAI to apply the techniques, using code examples from before and after Python to show the transformation of PHOTONAI using the technique. PHOTONAI is suited to this task as a high-level Python API designed to simplify and accelerate ML model development. It offers a unified framework to access existing ML implementations and integrate user-designed algorithms. Ultimately, Cottman sings the praises of PyCharm, stating that it “enables a good hunk of our development pipeline to be automated without learning a CI/CD package.” Why this is important: Although the examples in this detailed article involve Python, they would offer better understanding of results, better testing, less bugs, and a lower maintenance cost, for any programming language. [Click here to see the full picture!]( [Instagram Turns to AI to Protect Children]( In brief: Instagram has announced new updates in order to keep young people safer on the platform, including making it more difficult for adults to direct message teenagers, encouraging private accounts, and safety notices. Many of the changes focus on direct messaging, where Instagram users can privately chat to other Facebook users, as well as increasing the use of ML and AI to detect the real ages of users on the platform. Among the new features include blocking adults from sending messages to people under 18 who they are not already following; the adult user will then receive a notification from Instagram that direct messaging is not available. Instagram said: “This feature relies on our work to predict peoples’ ages using machine learning technology, and the age people give us when they sign up. As we move to end-to-end encryption, we’re investing in features that protect privacy and keep people safe.” Why this is important: Instagram say that data that can tell if an adult is sending a large number of friend or message requests to people under the age of 18, will be gathered and analysed by AI. Instagram has historically struggled to stop sexual abusers using the platform to contact minors. In the last two years, Facebook-owned apps (Facebook, Messenger, Instagram, WhatsApp) and Snapchat were used in more than 70% of sexual grooming crimes, and Instagram specifically was used in more than a quarter of cases. [Click here to find out more!]( [SuperDataScience podcast]( In this week's [SuperDataScience Podcast](, Stephen Welch joins us to discuss his 10 questions and learnings from 2020 on where machine learning can affect change in the world. --------------------------------------------------------------- What is the Data Science Insider? This email is a briefing of the week's most disruptive, interesting, and useful resources curated by the SuperDataScience team for Data Scientists who want to take their careers to the next level. Want more conversations like this? Earlier this year, we held our first-ever DSGO Virtual Conferences, where more than 3,500 data scientists gathered to learn, grow, and connect! If you missed them or want to repeat this fantastic experience, stay tuned to our upcoming virtual and in-person events that will take your DS career to the next level. DSGO is your go-to place to elevate your technical skills, gain life-long career lessons from industry experts, and build lasting connections with data-driven peers. If you want to learn more and register for our future events, [click here](. Know someone who would benefit from getting The Data Science Insider? Send them [this link to sign up.]( # # If you wish to stop receiving our emails or change your subscription options, please [Manage Your Subscription]( SuperDataScience Pty Ltd (ABN 91 617 928) 131, 63 Blamey, St., Kelvin Grove, QLD 4059, Australia