Machine learning fundamentally helps to solve problems that require pattern recognition. Most problems can be divided into the following six types
- when users have trouble finding what they need from a lot of information
Google and other searchers are a good example. Behind the scenes, we’re using a number of machine learning algorithms. It can also be used to automatically categorize more information. So, for example, this is an article about politics, this is a story about using Git in technology, and so on.
- when you want to recognize complex things
Automated driving has to be aware of the complexities that are around it. If it’s a photography service, it needs to recognize those images and automatically extract places, people, and other information from them. Such complex recognition can be achieved by feeding large amounts of data to pattern recognition algorithms.
- if you want to predict and forecast
Machine learning is useful when you want to make predictions, such as whether users will like the articles they see or not, or whether users will cancel their subscriptions. It can also be used when you want to do some forcasting, such as what your sales will be in the next three months, what your inventory will be in the future, and what your users’ access will be by region.
If you want to detect an abnormal value
It can be used to detect something that is moving out of the ordinary. Because machine learning is good at pattern recognition, it can detect anomalous behavior patterns that do not match what is recognized as a common pattern. This could be a fraudulent use of a credit card or an intrusion into an internal system.
- if you want to help with decision making
This is how Amazon and Netflix work when they recommend a book or movie based on patterns in the data of the user’s buying and viewing experience up to that point.
- if you need to communicate with a human being using words
If you’re building a service that communicates with humans, you need to use machine learning to do natural language processing, and assistant technologies like Alexa, Siri, and Google Assitant are exactly this case, translating human words into actionable tasks.
- if you want to create a new experience using augmented reality, etc.
For example, SnapChat offers a feature called the SnapChat filter that uses a facial recognition algorithm to creatively modify your face, which can be useful when creating new user experiences that have never been done before.