Remember when I tested the algorithm that recognizes cats and dogs in the pictures? No? Never mind. You can also test a much more advanced algorithm than Google that recognizes what is in the photo. How? Just download the Lens app from Google Play and enjoy the possibilities of Machine Learning. In today’s post, I would like to show you how Machine Learning is used today and when you can have contact with it.
Machine Learning in Autonomous Cars
Absolutely the most media application of ML algorithms is autonomous cars. You get in, choose the address to which the car should take you and the rest happens by itself. Pure magic. The source of this magic is the self-learning algorithm implemented in the car. The algorithm in real mode interprets the image flowing from the cameras installed in the car and, depending on the situation on the road, takes the learned action. Most likely, in the near future, a person will stop driving vehicles at all.
Already today, in addition to autonomous cars, we have autonomous trucks and ships. Personally, it is surprising to me that there is no information about autonomous trains. In my opinion, this is an area where the implementation of autonomous locomotives should not be a problem, but I probably do not know about it. In any case, man is in most cases the cause of road accidents, so if we want to increase safety in transport, we will not avoid eliminating human errors through full automation.
Captcha and Machine Learning
An interesting fact that I have recently met is the fact that in popular and disliked tests verifying whether you are a bot or a human (Captcha), ML algorithms are less likely to be wrong than humans. You already know that ML algorithms can recognize images. In practice, they do it much better than people. Probably for the same reason, ML began to be used to diagnose X-rays for early detection of lung cancer.
According to information made public, algorithms capture cancerous changes at a very early stage, before the human eye is able to catch any element in the photo indicating that something is happening. It’s also said that the future of diagnostics lies in dedicated ML algorithms. It’s easy to imagine why. One doctor can have only his knowledge and experience. One ML algorithm can have the knowledge of all the doctors of the world and all the knowledge written in the form of textbooks and even video and audio recordings. This enormously increases the diagnostic possibilities. Returning to X-ray examinations, here you will find more information on this topic: https://aijs.rocks/inspire/chest-xray-diagnosis/
The labor market is another case where Machine Learning is used
The third example of the use of ML algorithms, which I would like to present to you, is a virtual assistant in the labor market. It is called Emplobot and is the work of Polish programmers. The idea is that chatbot on the basis of a short conversation with you creates your virtual profile, and then adapts it to the needs of employers, collected from employers. The algorithm is constantly “learning” to better match the needs of employees and employers. Thanks to the work of Emplobot, employees, and employers save time. The former share their data for the first time and do not have to send hundreds of CVs to different companies. Employers, in turn, receive proposals from candidates after pre-selection. I like the idea very much and I support this initiative with all my heart.
Other areas of application of ML
I mentioned only three areas where ML is used, but there are many more applications of areas where ML works great. Google Assistant, e-commerce platforms, portals with legal advice, or even Windows 10. There are a lot of applications for ML and it is difficult for me to imagine the world in ten years, knowing how quickly the entire Machine Learning industry is developing and how widely it can be used. I only hope that everything that can be created will be used for the benefit of humanity.