When to use Machine Learning?

Machine learning is an approach to learn complex patterns from existing data and use these patterns to make predictions on unseen data.

1. Systems should have the capacity to learn.
2. There are complex patterns to learn.
3. Data is either available or can be collected.
4. The problem needs predictions.
5. Unseen data shares patterns with training data.
6. Problems are repetitive.
7. Wrong predictions have cheap consequences.
8. The solutions are at scale.
9. Constantly changing patterns.

• Don’t dismiss new technology for not being as cost-effective as existing solutions.
  • Technologies can become better with more investment.
  • If you wait for a technology to mature before adopting, you will be outdated.