LEARNER: Bias in Machine Learning

LEARNER: Bias in Machine Learning

logo

Inclusive Design Research Centre

wecount@inclusivedesign.ca

The Inclusive Design Research Centre is an international community of open source developers, designers, researchers, educators and co-designers who work together to proactively ensure that emerging technology and practices are designed inclusively. Learn more at idrc.ocadu.ca

Tags:

Artificial-Intelligence, Bias, EDI-principles, Machine-Learning, WeCount, Word-Embedding
The earner of this badge has learned how predictive algorithms and data mining affect different populations in a discriminatory manner. In addition, the earner has explored how specific data resources are used to train and reinforce machine learning models to produce biased outputs. The learner viewed a webinar on data bias delivered by Dr. Toon Calders, specialist in machine learning from University of Antwerp. A summative assessment was completed and then evaluated by a qualified WeCount assessor. Estimated learning time including assessment: 3 hours.

For more information, please visit We Count’s Home Page.


Instructions for Badge Earners

This assessment includes several multiple-choice questions based on the content of Dr. Toon Calders’s webinar on Machine Learning Bias. The second half of the assessment includes a reflection and short answer component to connect what you have learned to We Count initiatives.

This assessment is not time-restricted, featuring an option at the bottom of the form for it to be saved and continued later. For the multiple-choice, please be aware that CanCred selects the first option for each question to be the default answer. Remember to read through all of the questions carefully, and to take your time.

The average learning time for this badge is 3 hours.


Part One: Multiple Choice

Review Webinar if needed.
Review Webinar if needed.
Review Webinar if needed.
Review Webinar if needed.

Part Two: Short Answer and Reflection

Short paragraph to complete (Max 100 words). We encourage you to share your own examples of Data Bias. Alternatively, feel free to review or expand on examples from the Webinar if needed.
Put in the name of your example.
Please select one of the options from this scale.
Why do you think it is important to address this bias in the immediate future? This section is optional.
- or -