Dear Students,
You have done a great job learning how to take a business problem and converting it to a problem statement. That is typically the starting point for most AI courses out there. I am proud of how you have navigated your team discussions to unravel the business problem to find the pain of the customer and drive strategy discussions about several use cases possible to make a decision on what is actionable to solve the business problem.
You did not lose track of your business goals and set your own metrics of success for this AI project.
You looked at the data needs and how it ties to the potential type of AI application you were building and adapted your data and refined your problem statement.
All in 6 weeks live classes with global teams across time zones! Woot!
Now you are are ready for the Online Lab portion of our Capstone Lab course!
Remember our lesson about AI’s Confusion Matrix and definition of Recall and Accuracy. We will be using that information in next class. Here’s a quick reference from our class slides.
All AI is predictive and the AI Confusion Matrix shows the probabilistic nature of how the answers can be positive or negative. This does not tell you how often does the model predict positive and negative results.
That is where the terms “Recall” and “Accuracy” come in. Recall tells you how often the model predicts positive results. And Accuracy inform how accurate is the model in predicting the correct results.
Each of your journeys in your multi-disciplinary teams to learn a data science sprint to build an AI for a real business problem is unique from your past business experiences and functional job lens.
Product Managers look at it as AI Product Management and I heard about an AI PRD being developed. Business Managers look at it from business model development and what data partnerships will help the business. There are many directions you can take this strategically by focusing on internal data and prepping it or adding external data sources from partners to enrich your use case. Thats the power of possibilities with AI!
This week, we will setup the online lab, again, remember our promise to do this with NoCodeAI. The class is moving to the weekend (Sat) and we will spend 2 hours on the online lab.
We will look at IBM Watson Service using IBM’s AutoAI on IBM Cloud. We will also compare with setting up the same on Akk.io’s AutoML platform.
In the first hour we’ll learn 5 classification models. No, you don’t need Python or Linear algebra for my class.
In the second hour, we’ll setup the lab and fit a simple binary classification model.
You can take this learning and go fit your industry partner’s data as next step.
Excited to see you at our online lab on Saturday!