Machine learning applications - healthcare, retail, transportation and more
We started this series of articles with a general article in which we outline the links between machine learning, neural networks and deep learning. These last two topics will be discussed later. In these articles we first discuss machine learning. In the previous article we gave a definition of machine learning. Subsequently we now look at the applications of machine learning. Companies are increasingly realizing the incredible potential of unraveling the wealth of information that IoT and Big Data can deliver. That is why they are increasingly applying machine learning. We discuss machine learning applications for retail, transport, image recognition (in healthcare), and entertainment such as games.
Applications of machine learning - retail
Everyone must have experienced it before: you buy a certain product and you immediately get recommendations for other, comparable products. This is almost the most common form of machine learning that you now see in practice. Many large web shops collect data from visitors and store this data. This stored data allows web shops to respond to the interest of potential customers. In addition, machine learning keeps prices as low as possible by continuously comparing them with other retailers and machine learning is used to filter out unusable product reviews. A second example is fraud detection, which determines credit risks. Via machine learning, relationships in data can be traced to determine whether a transaction is fraudulent.
Applications of machine learning - transport
Machine learning is also used in the development of software for autonomous cars. Hereby all images of traffic situations are entered into a system. Ultimately, the software for autonomous cars must make its own decisions about whether to brake for certain situations or not. Machine learning ensures that an autonomous car knows the difference between a pedestrian and a traffic light, and for example knows that the speed must be adjusted if it rains hard. In addition, manufacturers are developing systems that can predict when a car breaks down. This gives a huge number of advantages with regard to logistical planning of, for example, trucks. When there is a long drive ahead, it is sometimes better not to send a particular truck on the road. want see institute for Machine Learning Course in India.
Applications of machine learning - image recognition
Machine learning allows computers to train to recognize and name specific images. Anyone with a Google Photo account can test that themselves: enter a subject such as 'flowers' or 'dogs' and watch Google take pictures of your petunias or your golden retriever. You never tagged those photos as such, so how can Google know that the animal in the photo is a dog, not a cat or wolf? Another example of machine learning is DeepArt. This model is trained in the painting style of Vincent van Gogh and can transform photos into an image in the style of the painter. There is no record in DeepArt of exactly what that style is, but the program has learned it by training itself on a huge dataset of paintings.
Applications of machine learning - healthcare
Applications of machine learning - games and entertainment
Because a machine learning model is trained on the basis of data and therefore has no fixed instructions, a model is not limited to predefined situations. This enabled AlphaGo, for example, to beat the best Go player in the world. It was not necessary to write strategies for the program in advance: these were developed by themselves during playing. DeepMind also learned to play dozens of Atari games on its own. Facebook is also no longer possible without machine learning. Algorithms help the social network, for example, to recognize faces, personalize the News Feed, rank search results and remove hurtful content. A system called DeepText was recently added. This algorithm can 'understand' posts and messages, according to Facebook almost as good as people. Are you telling a friend in Facebook Messenger that you need a ride? Thanks to DeepText you will immediately receive a link to a taxi. Finally, machine learning is used to connect online gamers who are matched in terms of level.
See also: Tips to best learn a particular programming language/
