If you are here, then either you are curious about what machine learning is or explore and learn about machine learning. No worries, everyone is a beginner at their initial point. Here I have shared what is machine Learning and the basics of Machine Learning in easy and understandable way.
What is Machine Learning?
Machine learning is a subfield of Artificial Intelligence. The main goal of machine learning is to understand the structure of our data and fit that data into our models which could be understood and utilized by us. Machine learning is unlike the traditional computational approaches because, in the traditional computational approach, the problems are solved using the sets of explicitly(long and detailed) programmed instructions. But Machine Learning the model is built by training the data inputs and use statistical analysis to find the output values which fall within a specific range. The models are built using sample data from the inputs to automate the decision-making process.
Machine Learning is defined with two definitions from two different people.
A person named “Arthur Samuel” in 1950 described Machine Learning as:
“The field of study that gives computers the ability to learn without being explicitly programmed”.
This means that when a computer is able to make decisions or act accordingly without writing code in a very detailed way. This is defined as an older, informal definition way of defining Machine Learning.
Another person named “Tom Mitchell” on 1998 provides a modern definition for Machine Learning stating that :
“A computer program is said to learn from experience E with respect to some class of tasks T and performance measure P, if its performance at tasks in T, as measured by P, improves with experience E.”
E = the experience of playing many games of checkers
T = the task of playing checkers.
P = the probability that the program will win the next game.
This means the computer will learn from its experience based upon its performance on the tasks which will literally improve the experience of the computer
Types of Machine Learning
1) Supervised Learning Algorithm
The term “Supervised Learning” refers that the dataset will be given with actual output and the algorithm will learn from it to predict and identify the output by finding the relationship between the input and the output. Supervised learning problems are categorized into 2 types:
- · Regression — It helps us to predict a continuous outcome variable based on the value of one or multiple predictor variables and uses the correlation between the features to predict the continuous output.
- · Classification — It is a process of categorizing a set of data into classes. It generally groups multiple classes based on the input features and categorizes the data to a particular class based upon the similarity of the data with respect to the class.
Examples of Supervised learning:
- House Price Prediction
- Breast Cancer Prediction
2) Unsupervised Learning Algorithm
In Supervised Learning Algorithm, we are given a dataset without proper labels in it and we have to approach the problems with little or no idea of what our results look like find the structure with the help of the data’s where we don’t know the effect of the variables in the dataset properly. Types of unsupervised data:
- Clustering — It is a task performed to segregate groups with the help of given data with similar features. So that we could identify the groups where the data belongs to even without proper labels of the data.
- Dimensionality Reduction — It helps us to reduce the dimension of the data by removing unwanted features and the features which are having higher correlation or having low variance.
Examples of Unsupervised learning:
- Advertising Platform
- Customer Segmentation
3) Semi Supervised Learning
When a dataset is present with both labelled and unlabelled data where, usually small amount of labelled data and large amount of unlabelled data will be present. This method is used only when availability of data’s for training the dataset is minimum. The information generated ny the machine during this partial training is called as pseudo data. These are most commonly used in medical fields.
Examples of Semi Supervised Learning
- Protein Sequence classification
- Speech Analysis
4) Reinforcement learning
It is the training of machine learning models to make a sequence of decisions. It helps in solving complex problems and helps to take suitable action to maximize reward in a particular situation with the best time complexity. Reinforcement learning is widely used in Artificial Intelligence and robotics
Examples of Reinforcement learning
- Self-Driving Cars
Where are we using Machine Learning in our day-to-day life?
- Email Spamming — We are getting flooded with email in our mailbox and most of the mails are spams(unsolicited and unwanted junk) email which are used for commercial purpose. Here Machine Learning plays an important role by classifying them into spam or not.
- Product Recommendations — Almost many of us are shopping in E-commerce websites. have you ever wondered how do you get recommendations? Yes, they use Machine Learning techniques to find a similar product which you are searching.
Hope you had understood what is machine learning and what are the types of Machine Learning and where are they used in our real life.