It’s not long ago that AI was a semi-mystical concept that belonged in science fiction films. But although AI and machine learning are moving into the mainstream, many IT leaders see machine learning for IT operations as an interesting but complex topic. Doesn’t machine learning mean lengthy engagements with expensive data scientists? In a word, no. You can apply machine learning to mainframe data to help you make IT Operations decisions that are better for your business. And doing it’s a lot more straightforward than you might think.
First, let’s pin down a few key terms. Forrester draws a helpful distinction between “pure AI” that strives to mimic human intelligence and “pragmatic AI” that applies a moderate level of intelligence to applications. Machine learning is an example of pragmatic AI in action. In machine learning, algorithms analyze data to find models that can predict outcomes or understand context with significant accuracy.
Learn and predict
So how does machine learning enable predictive analytics? These intelligent algorithms learn patterns in data and use what they learn to predict similar patterns in new data. Machine learning automates this process, equipping computers to make progressively more accurate predictions. In real time. Without human intervention.
Intelligence in action
Let’s take a real-world example of machine learning in action: a smart thermostat. With a regular thermostat, you simply turn it up and down as the temperature and seasons change. But with a smart version that incorporates machine learning, your thermostat gets to know your preferences based on your behavior and applies them autonomously.
This learning process is based on patterns of behavior rather than one-off instances. Your smart thermostat will learn your preferred settings for different times of day across different rooms, based on whether it’s a weekend or a weekday, and predict the temperature you want with ever-increasing accuracy. Even smarter, these devices connect to Wi-Fi so they can factor-in weather pattern changes and forecasts, and react dynamically to the outside environment. The results? Lower energy bills and more convenience to take just two examples.
Now the really good news. Machine learning and predictive analytics are becoming available within mainstream IT operations management tools. So you don’t need a data scientist painstakingly building and refining algorithms to enjoy these advantages. Next time, I’ll take a closer look at where machine learning and predictive analytics help your run a more efficient and highly available mainframe.