Human intelligence meets machine intelligence
Machine Learning (ML) has transformed the art and science of business. Its algorithms provide a powerful analytics capability that organizations across every sector are keen to harness. But be warned. ML is not a magical crystal ball that automatically produces simple, clear-cut answers to questions about the future. To be successful, it requires skilled handlers who understand the right algorithms for the job, while being clear about its limitations. Shortcomings may lie in the underlying data, the particular algorithm used, or how well we can explain the very predictions we want ML to deliver. There is an inherent tension between explainability and predictability; more explainable techniques are usually less accurate. But more accurate techniques are far less transparent, and therein lies the tradeoff and the critical question: how comfortable do you feel about relying on the predictions of ML when the results are difficult to explain?
The Evite Machine Learning Data Analytics Application exposes learners to all these areas of ML, from the technical aspects of data wrangling and running ML algorithms to analyzing the outcomes and bringing the insights back into the business decision-making process.
This data application has three approaches instructors can select from to match the desired coding complexity to the technical capabilities of their learner audience. In all three options, learners will gain access to Jupyter and will be exposed to the analytical tools:
Strategy: for learners who have little to no coding or first-hand data analytics experience, but are keen to gain an understanding of how ML works. This approach is suitable for non-computer science classes, executive & manager audiences.
Hybrid: for learners who have moderate coding skills, and may or may not have first-hand data analytics experience. This approach is suitable for learners who are looking to take their existing technical skills to the next level.
Complex: for learners who are experienced Python coders, and may or may not have first-hand data analytics experience. This approach is suitable for computer science classes, and adult learners who are in the technical field.
Available Fall, 2021
Company: Wharton Interactive
Approximate price: Unknown
Playtime: 4-6 hours
Learning objectives: ARC delivers the business problems the students must solve, promoting strategic decision-making skills rooted in a finely tuned analytical mindset. As the ARC narrative progresses, the learner's Jupyter notebook unveils deepening complexity, delivering the case related data, analytical tools (Python and AI/ML algorithms), and varying levels of code to solve the business problem.
Link to simulation home page: https://interactive.wharton.upenn.edu/experiences/evite-machine-learning-data-application/