Machine learning is all around us. It is how Netflix is able to suggest the next movie for you to watch. It is why ads on the internet seem to know what you are interested in. It is also how autocompletion is accurately able to predict the next word you might type.
But what exactly is machine learning, and how does it work?
Simply put, machine learning is a subset of artificial intelligence that involves the development of computer algorithms that access large amounts of data to create models for information. These models are then used to predict specific behavior. For example, if someone tends to watch a lot of movies involving racing and cars, a good recommendation might be a film on the history of cars or car restoration.
Sponsored by Digi-Key
Knowles DLI High Frequency 5G SMD Filters are Small Size Fully Shielded Components Available from Digi-Key
The 26/28/39 GHz catalog filters provide 3 GHz of bandwidth, >50 dB rejection, are 20x smaller than others while implemented in surface mount packages for standard tune-free assembly, and provide temperature-stable operation from -55°C to +125°C.
Machine learning consists of several key steps. The first step is gathering and sorting data, and then developing a model based on the data. This model is then trained, evaluated, tuned, and then used for predictions. As more data is collected and more predictions are made, the machine or algorithm continues to learn and improve its predictions.
The more data the machine has to base its predictions off of, the higher the probability that the the prediction is accurate. Thus, machine learning is the process of inferring or deducing new information based on existing data, and typically lots of it.
Machine learning is not a new concept. In fact, the term “machine learning” was coined in 1959 by Arthur Samuel partially based off of a checkers playing program. The program would choose the best next move to minimize the potential to lose the game. Due to the limitations of the computational hardware of the time, Samuel had to use a method of deriving the answers called alpha-beta pruning.
Machine learning today is much more technologically advanced. Computational resources are more powerful–and cheaper and more accessible. Computer memory also is larger, faster, and cheaper. As a result, machine learning algorithms are able to crunch complex mathematical equations incredibly fast.
Any industry that deals with large amounts of data is effectively using machine learning to gain insights and knowledge. The financial industry uses machine learning to help identify and prevent fraud. Google’s self-driving car is a technological innovation built off of machine learning. Many healthcare organizations are now using machine learning, along with various sensors and wearable devices to help identify potential issues and trends relating to medical conditions.
Machine learning will be a core part of technological advances, such as artificial intelligence and the internet of things, for the foreseeable future.
Machine learning helps create “autonomous” building
NXP partners on Arm machine learning accelerator