Machine learning is the code of much futuristic technological advancement in today’s world. We can see many application of it around us such as tesla, self driving car, apple series and many more. This is the subset of artificial Intelligence (AI) and mainly focuses on the designing of the systems and thereby allowing them to learn and make predictions based on some experiences which is data in case of machines. It allows the computer to act and enable the data driven decisions rather than explicitly program to carry out a certain task. These are implemented to improve our time when exposed to new data.
Supervised learning: This is the process of each instance of training data sets which is composed of different input attributes and expected value of output. The input attributes of data sets can be pixel of image, value of database and so on. For each input instance there will an expected output is associated. The algorithm learns the input pattern and generates the output pattern. Some popular examples of supervised learning is the simplest one of making a phone call and getting it forwarded to the concerned calling number. Any speech automation in the mobile is trained with the voice. Once it is trained it works based on the training. Weather application, Biometric identity and many more can be done. In banking sector the supervised learning recognizes the credit worthiness of credit card holder. In health care sector it is used to recognize the patient readmission rates. In retail model it is used to analyze the products customers buy together.
Unsupervised learning: This is the process of each instance of training data sets don’t have an expected output associated to them. Instead the algorithm detects pattern based on the in it characteristics of the input data. Clustering is the example of unsupervised learning. This model can recognize the similar data types into the cluster. For example a friend invites you to the party and where you meet totally strangers , then you classify them based on unsupervised learning as you don’t have any idea about them and this classification can be done based there gender, dress, education or whatever way you might like. This learning is different from supervised learning because you don’t have any prior knowledge about the people so kept on classifying them as they kept on coming to you. This learning in banking sector segments customers by behavior characteristics by serving prospects, and develop multiple segments using clusters. In health care by categorize MRI data by normal or abnormal. In retail sector it is used to recommend products to customers asked on past purchases.
Reinforcement Learning: The machine learning tools algorithm allows the machines and software agents to find out the ideal behavior with the explicit context to get the most out of its performance. The reinforcement learning in banking sector create ‘next best offer model for the call center’. In health care in allocates scarce medical resources to handle different ER cases. In retail sector it reduce excess stock with dynamic pricing.