A lot of people are confused about what type of use cases Machine learning and AI is good for?
Yes everyone knows Machine Learning projects require data but so are other analytics projects. So the question becomes where machine learning fits in our overall analytics landscape? For what types of problems, it’s suitable?
Before I answer the above question, let me list down the types of analytical projects that any enterprise might be doing, then we will discuss where Machine learning fits in.
Analytics projects fall into four major categories
- Descriptive analytics: How is the business performing?
- Diagnostic Analytics: What caused the sales decline?
- Predictive Analytics: What is the sales forecast for next quarter?
- Prescriptive Analytics: What is the next best action?
Diagnostic Analytics
It is the Analysis that helps to diagnose issues and root cause
Used to answer questions such as…
- What caused a decline in sales?
- Why did a region miss its target?
Predictive Analytics
Forward-looking analysis to anticipate the future
Helps to answer questions such as…
- What is our sales forecast for next quarter?
- Which customers are likely to default?
- Which prospects are most likely to buy our product?
Prescriptive Analytics
Gives a clear recommendation on the best course of action
It helps to answer questions such as…
- What is the next best action?
- How should we invest our money?
- What is the next best move? (AlphaGo)
- What is the best route to my destination?
As you might have guessed from the type of questions
Traditional BI and reporting do descriptive and diagnostic analytics. While Machine Learning and AI thrive on predictive and prescriptive analytics
This understanding will help you solve select problems that are suitable for AI. As any in any field, AI projects have a specific agile and iterative journey to make sure the investment is value-driven.
Let me know in the comments about your feedback or questions.