Data Science Digest 10

As a data professional, time is at a premium. Here are some tips and trends you’ll want to stay on top of!

Title: Towards open health analytics: our guide to sharing code safely on GitHub

Source: https://towardsdatascience.com/towards-open-health-analytics-our-guide-to-sharing-code-safely-on-github-5d1e018897cb
Author: Fiona Grimm
How: Provides step-by-step instructions and things to consider
When to use this: When preparing to create a GitHub page, especially which may include sensitive data
Why it’s helpful: Case study with tips, instructions, checklist and links from someone who has done this before
Suggested application: Contribute to and benefit from the input of the global community
Business impact or insights to be gained: Good reference to provide management who might be resistant or concerned about sharing company code or information

Title: SDC Handbook (Safe Data Access – Statistical Disclosure Control)

Source: https://securedatagroup.org/sdc-handbook/
Authors: Emily Griffiths (University of Manchester), Carlotta Greci (The Health Foundation), Yannis Kotrotsios (Cancer Research UK), Simon Parker (Cancer Research UK), James Scott (UK Data Archive, University of Essex), Richard Welpton (The Health Foundation), Arne Wolters (The Health Foundation), Christine Woods (UK Data Archive, University of Essex)
How: This 3-part handbook covers what SDC is about, why it matters, advices on assessing specific statistical outputs and how to set up SDC systems and user requests. The approach is built upon the “Five Safes” – Safe Data, Safe People, Safe Projects, Safe Settings, Safe Outputs
When to use this: When you need to ensure that “statistical results produced from confidential data pose a minimal risk of disclosure of identity and/or personal information.”
Why it’s helpful: there aren’t many resources available that walk through how to do this on a day-to-day basis
Suggested application: Government, academia and commercial settings, such as healthcare
Business impact or insights to be gained: practical steps for assessing risk, confidence for data owners that access to data is securely managed and confidentiality will not be compromised, facilitate the process of requesting releases of statistical results, and guidance for training of staff working with data where “Safe Settings” is employed.

Title: DoD Adopts Ethical Principles for Artificial Intelligence Use

Source: https://healthitanalytics.com/news/dod-adopts-ethical-principles-for-artificial-intelligence-use
Author: Jessica Kent
How: “The DoD Joint Artificial Intelligence Center (JAIC) will coordinate the implementation of ethical AI principles for the department”
When to use this:”DoD’s ethical principles include several major areas: Responsibility, which ensures that DoD personnel exercise appropriate levels of judgment and care while maintaining responsibility for the development and use of AI; equitability, which will confirm that the department will take steps to reduce unintended bias in AI algorithms; and traceability, which will ensure that the relevant personnel have an appropriate understanding of AI capabilities.

“Additionally, the principles will ensure that the department’s AI capabilities will have explicit, well-defined uses, and that AI is designed to fulfill intended functions while detecting and avoiding unintended consequences”

Why it’s helpful: If seeking military contracts, be aware of these requirements up front/ make changes now
Suggested application: AI and machine learning development
Business impact or insights to be gained: Combat and non-combat functions are subject to these principles

Be sure to subscribe to this blog for more tips. And sign up for our free tools to learn of job opportunities and more.