Machine Learning and its Algorithms

Machine learning is a branch of artificial intelligence, study to develop programs that automatically learn from data without explicitly programmed. How a machine learning program can be programmed without programming, algorithms are used in machine learning. Machine learning algorithms are mathematical functions that model and predict data. These algorithms are used to train from data.  The computer programmers who train and test algorithms and deploy on the platforms are called machine learning engineers.  Machine learning has also created new jobs such as machine learning engineers, Machine learning researchers, and machine learning algorithm designers. ML engineers mostly work in software companies, ML researchers work in academia and Research and Development (R&D) labs in industry, Machine learning algorithms designers work in industry and academia.

Machine learning is divided into three types.

SUPERVISED LEARNING          

Supervised learning s type of machine learning where algorithms train with label data. The label is a category in the dataset which the algorithm trained with a given label, the algorithm learns data mapping from given label data. Let an example of supervised learning. For example, a company needs a system that recognizes cars. It is called an object detection problem. Machine learning algorithm will be trained with thousand of images dataset contains labels or classes Car, Not Car. So algorithm after training will recognize Car or Not.  This is called Classification and this problem is binary classification problem because algorithm train and predict two classes. Linear regressions, Logistic regression, Support vector machine, decision trees, k-nearest neighbor are examples of supervised machine learning examples.

UNSUPERVISED LEARNING     

Unsupervised learning s type of machine learning where algorithms need no labeled dataset. Unsupervised learning algorithms find patterns in data and predict according to discovered patterns. Clustering is an example of unsupervised machine learning. Let an example of supervised learning. For example, a news agency does not want to hire more people who manually work to group similar articles, so the agency has hundreds of articles and needs a software which groups the similar news such as international news, technology news. The clustering algorithm will be trained with news articles corpus or dataset and will make a cluster of news which have similarity. The clustering will group similar articles in clusters and will provides list of clusters or categories of similar news which the news agency will publish on their digital platform. K-Means, hierarchical clustering, DBSAN clustering, OPTICS algorithm, Autoencoders are examples of unsupervised machine learning algorithms.

 REINFORCEMENT LEARNING

Reinforcement learning s type of machine learning where algorithms learn from the environment and take action to get the maximum reward and minimum penalty. Reinforcement learning does not label data so it learns from experience and maximum rewards. Q-learning algorithm is an example of reinforcement learning algorithm. Reinforcement learning has many applications, it is mainly used in games.

Tags: Machine Learning, Machine Learning algorithms, Supervised Learning, Unsupervised Learning, reinforcement learning