MACHINE LEARNING FOR DECISION MAKERS
CHAPTER

2

Understanding Machine Learning?

An overview of practical examples and applications in every industry

David Mora
|
5 MIN READING TIME

Machine learning is simply the process by which a computer learns patterns from historical data, and then uses those patterns to make predictions about the future.

Here's a practical example from real estate:

1. Given historical real estate data, a computer learns the relationship between home selling price and data about that house: for example the number of bedrooms, square footage, and location. Then a computer might learn that the price of a house increases by $10,000 for every bedroom it has.
2. The computer can now accept data about a house that hasn't sold yet, and accurately predict its selling price using those data relationships.

In Machine Learning, computers learn the “rules” hidden in the data that predict a key business outcome. (This is unlike software, where the “rules” computers follow must first be written by programmers.)

Machine learning has powerful applications in every industry:

  • Service providers use data on past customers to predict which customers are likely to stop using them, and then prevent it;
  • Sales teams use data from past leads to predict which new leads are most likely to buy their product;
  • A school can use data on past students to determine which new students are most likely to drop out, and then intervene proactively; and
  • A warehouse can use data on past equipment logs to predict which machines are most likely to breakdown in the future.


With large datasets, machine learning does what humans can’t:

  • Process massive amounts of information in seconds;
  • Detect subtle relationships a human would have overlooked;
  • Give precise explanations of which data fields most predict your key outcome; and
  • Quantify how certain a prediction is.