What is machine learning?
Machine learning is a type of, ai aka artificial intelligence that empowers a system to learn and make decisions itself without being programmed. These algorithms make the computer smart enough so that it can make choices on the basis of the data it has without any human intervention. The primary aim is to make algorithms that allow a system to learn and make their own decisions in future, based on the past data two why do we need machine learning.
Given below are some of the reasons we use these in the here and now.
Prediction while traveling
we all have been using gps system while traveling in our lives. whenever you book a cab it tells you the approximated fare and time required to reach your destination.
How does your smartphone do that ?
The answer is machine learning it calculates the velocities and location of our vehicles. Based on this information, it even tells us if there is traffic jam on this road. The programmers did not program the computer to tell you that there is a traffic jam, but they design a system that makes smart decisions on the basis of past and current events of people who pass by that area. Plus, it warns you about the traffic jam.
Search Engine Optimization
web search engines automatically show you the accurate results based upon your location in past searches. Programmers don’t program it to show you those results, but it gives accurate results within seconds according to your interests in recent searches
Span Mail Classification
in our email boxes, the system automatically classifies some emails as spam or junk mails and some mails as primary mails that could be very important for us. The system is never wrong and it is all possible with the help of these learning.
Three types of machine learning the basic idea of machine learning is the same for all types but it has been further divided into three following types:
Supervised Machine Learning
learning is one of the most popular types of machine learning and it is easy to understand and implement. In this type, the algorithm is trained on given data but in the data needs to be labeled. You allow the system to predict the data and you make corrections if the predictions it makes are not accurate enough.
Unsupervised Machine Learning
unsupervised machine learning works without any labeled data but you have to provide a lot of data so that the system understands the properties that provide a base for the decision it has to make. This can improve the productivity in a lot of fields.
it is based upon trial and error methods. The system makes mistakes and learns from them in order to avoid these mistakes again. For example, in a maze, when the system fails to find a path it won’t go on the same path again because it knows that the path doesn’t work. It labels positive outcomes and negative outcomes and runs on the basis of these outcomes.
In short these were some of the common questions about machine learning hopefully the answers to these questions will help you get a deeper insight into this field of science