What is Machine Learning & Types of Machine Learning

What is Machine Learning?

As the name suggests, machine learning is something that is teaching a machine or a machine is learning something. This is nothing but a computer program that can learn on its own by using data.
According to Arthur Samuel (1959), Machine learning is the field of study that gives computers the ability to learn without being explicitly programmed.

A more specific & engineering-oriented definition is, A computer program is said to learn from experience E with respect to some task T and some performance measure P, if its performance on T, as measured by P, improves with experience E.
This definition is given by American computer scientist Tom Mitchell in 1997

 

Since we’ve learned about what is Machine learning, now we need to know about the types of Machine learning Systems. There are many different types of Machine Learning, e.g

1. 1. Supervised Learning
   2. Unsupervised Learning
   3. Reinforcement Learning
   4. Semi-Supervised Learning

These are the four types of Machine learning systems that are often used. Remember sometimes you may listen to three types of learning systems excluding the last one. Let’s understand the different types of ML systems in detail.

Before understanding the types of systems let’s first about feature or label which is closely related to Machine Learning or ML. A label is a final outcome or the desired output that you want from a Machine Learning algorithm. This may also consider as the target.

 

Supervised Learning:

In this type of Machine learning the dataset that is fed to the algorithm is labeled or the data is a labeled data. Based on this label the algorithm learns & can predict the future outcome. For example, let you are running an old car store & now you got a huge delivery of old cars & you want to forecast the price of them based on the previous car’s price. Here the price of old cars is a label which the algorithm will use to predict the new car price & hence the data is feeding a label data this problem can be solved through supervised learning.

Some of the supervised learning algorithms are:

• Linear Regression
• Logistic Regression
• K-Nearest Neighbors
• Support Vector Machine (SVM)
• Decision Trees & Random Forest

 

Unsupervised Learning:

In this system, the data that the algorithm is taken is not labeled data. The dataset has no label to predict. The algorithm used in Unsupervised Learning used to find the pattern of the data & solve the given problem.

For example, let you have a blog that has huge content on multiple niches & now you’ve published two new posts on Machine Learning Algorithms & web development. So, you want to find the exact reader who is interested on machine learning and who is in web development for a marketing campaign and you don’t know who are those readers from different readers.  Unsupervised learning will help to make segments of the different readers in your blog. This type of ML algorithm will help to find patterns from the dataset & group different readers into different groups for an easy email marketing campaign.

Some of the commonly used algorithm in unsupervised learning is:

• Clustering
• K-means
• Density-Based Spatial Clustering of Application with Noise (DBSCAN)
• Hierarchical Cluster Analysis (HCA)
• Association Rule
• Apriori
• Eclat

 

Reinforcement Learning:

In this type of machine learning system, there is no dataset instead of the algorithm works as an agent. The system learns to form the environment. This is just like a toddler who learns from the environment. When the algorithm does something correct it gets a reward or if it does something wrong it gets a penalty.

For example, let’s there is an algorithm & on its left side there is fire & opposite side there is water. Whenever it goes closer to the fire it will burn or get penalty & next time it will find it dangerous & won’t go there again. & If it goes closer to water it will get a reward & it found the water not harmful for it.

 

Semi-Supervised Learning:

In the type of learning method, some data has labels & some data is unlabeled thus it is known as a semi-supervised machine learning system. This type of machine learning algorithm is a combination of supervised & unsupervised learning techniques.

Conclusion

These are the main four types of machine learning methods. There are some other types of learning those are according to the platform of learning & types of learning. Read from Here.

Hope you understand these four types of the learning system.  If there is any confusion or problem let’s know so that everyone can know it.