As we have discussed what machine learning means, let's find out the types of machine learning algorithms. Machine learning algorithms are classified into three types based on approach. Let's head to the first one.
1. Supervised Learning
In supervised machine learning, we supply the algorithm with training data sets and provide both inputs and outputs. The algorithm maps the input and output based on the training sets. Supervised Learning includes both classification and regression algorithms. The classification algorithm is used when the outputs are limited to a value. The regression algorithm is used when the outputs are not limited and are continuously varying.
2. Unsupervised Learning
In Unsupervised Learning, the training data sets are not provided. The machine is just provided with inputs and outputs and the algorithm is made to act without any information.
3. Reinforced Learning
In Reinforced Learning, there are three parts namely agent, reward and punishment. The agent has to reach a reward with less number of punishments. The reinforced learning is more like a trial and error method. The best example of reinforced learning is a Chess game. The input for every step taken.
Applications of Machine Learning
There are many applications ranging from the simple spam filtering to complex autonomous driving
Have you ever noticed that some of the quite important emails also being delivered to your spam box? This is because of the machine learning algorithm. It sometimes finds similarities between the important and spam due to which it gets targeted as spam. That is why every subscription asks you to check the spam folder to avoid further delays.
Data mining is done in order to improve the UX[User eXperience] and other research purposes. Data mining uses a machine learning algorithm to collect data and sort it as per requirements.
Machine learning is also widely applied in the automation of image processing.
Online streaming sites
Online streaming sites use machine learning algorithms in predicting the viewers' count and people's interests. They are always going through the comments section through the machine learning algorithm.
Autonomous Driving systems also learn from real-time experiences and work effectively.
You might have observed that whatever you were searching for on Google is found in the top three. It is because of the machine learning algorithm that ranks the search results.
Is Machine Learning harmful?
Yes, it is. Let us get to an example. In the year of 2014, Amazon started using a machine learning algorithm to recruit candidates. After a period of time, it found out that female applications were discriminated. This happened because in machine learning, the learning begins from humans and hence it faced that problem. As machine learning takes from humans, machine learning is considered risky.