The word ‘Machine Learning’ is not new again as it has been heard consistently. Machine Learning is transforming every industry and can be used in any domain to drive innovative solutions to businesses.
From Computer Science we understood and learned that programming is a set of instructions or tasks given to a computer to perform, these set of instructions can be from any language of choice but of course it will be the language the computer understands,
On the contrary, machine learning is the process in which we allow computers to learn from our data in order to find some patterns.
Arthur Samuel in 1959 defined machine learning as the field of study that gives computers the ability to learn without explicitly being programmed.
So, we can say machine learning is the study of computer algorithms and a part of artificial intelligence that improves automatically as a result of experiments and the use of data.
From Wikipedia, we can say machine learning algorithm builds models, in layman we can say machine learning algorithms build formulas based on sample data, which we can call training data in order to make predictions or decisions without being explicitly programmed.
In ordinary programming, the software developer or the programmer writes every single rule that makes up the program or task in he or she is trying to perform. But in machine learning, the data is fed into a machine learning model and the job of the model determines a set of rules that maps the data and the label.
So we can say traditional programming takes rules and data to produce results while machine learning takes data and results(labels) to produce rules.
Application of machine learning
As I inferred in the earlier sentence, the application of machine learning knows no bounds and the application can be into several domains.
- Churn Prediction: In machine learning, we can predict whosoever is likely to leave a company, organization, or system or not. In this case, the data and the result of historical people in the company will be fed into the machine to give us some rules and which are in turn used to predict new clients or customers.
- Movie Recommendation. This case study can be seen in youtube and Netflix video recommendations.
Though machine learning isn't suitable for all problems,but it is suitable for problems or programs that are too complex and those programs that it is impossible to write all rules for.
Machine learning can also be used heavily on some applications and with extra care with a human being involved, this involves some areas like self-driving cars.
Types of Machine Learning Systems
There are 5 main types of machine learning systems, these are:
- Supervised Learning: This is a type of machine learning in which the model is trained with ‘input data’ along with some form of guidance known as a ‘label’ or ‘target’ or ‘output’. In other words in order to carry out supervised learning, there must be input and output. Broadly, we have 2 main kinds of supervised machine learning problems, these are:
i.Classification Problems: In classification tasks, we are to identify a given ‘category’ from numerous categories.
ii. Regression Problems: The goal is to predict a ‘continuous’ value of a dataset. An example is to predict the price of a house based on some features.
2. Unsupervised Machine Learning: These types of machine learning are trained with ‘unlabelled data’. They are mainly used for clustering.
3. Semi-supervised Learning: This type of machine learning falls in between supervised and unsupervised learning. In semi-supervised learning, a small portion of the training data is labeled while the rest of the data points are not labeled.
4. Self-supervised learning: This machine learning uses the entire ‘unlabelled data’ and does not require any manual labeling of data.
5. Reinforcement learning: This is a special type of machine learning that is most applicable in robotics and games. In reinforcement learning, a system called an agent can perceive the environment, performs some actions, and get rewarded or penalized depending on how it is performing.
What is Artificial Intelligence, Data Science, Machine Learning, and Deep Learning?
Often time, artificial intelligence, data science, machine learning, and deep learning are used interchangeably but they are quite different.
Artificial intelligence is a branch of computer science and an interdisciplinary field concerned with building intelligent machines capable of performing tasks at the level of humans and mimic humans.
Machine learning is the branch of AI concerned with giving the machine the ability to learn from data.
Deep learning is a branch of machine learning that deals with the study of artificial neural networks and it was inspired by the human brain.
Data science is an interdisciplinary field that deals with using data to solve business problems using various techniques.
Reference:https://nyandwi.com/blog.html
