“What is machine learning?” It’s a question that gets asked a lot as organizations think about implementing machine learning-based solutions to help them stay ahead of the game. Understanding the fundamental principle behind machine learning (ML) can help you plan where to use it in your organization and use the right approach to building an ML system.
As you’ll see in this video, rather than hard-coded logic, ML systems are trained much the same way we are. The key to training the system is having enough up-front data to feed it. The system can then learn important attributes and make decisions. For example, an insurance company training an ML system on the differences between a low-risk and high-risk asset should feed the system multiple instances of each.
Testing the system is crucial. In our example, the company can test by presenting unlabeled assets and then verifying that the system can correctly identify which are low and high risk. Once trained, an ML system can potentially automate a manual process that is causing a bottleneck.
Watch more IBM Developer Diaries.