Question

I need new and unique answers, please. (Use your own words, don't copy and paste), Please Use your keyboard (Don't use handwriting) Thank you..

Supervised vs. Unsupervised vs. Semi-supervised Learning

Data scientists use many different kinds of machine learning algorithms to discover patterns in data. These algorithms can be classified in three main categories: supervised, unsupervised, and semi supervised learning.

For each Learning type, give an application and explain why we should use it?

Answer 1

Supervised Learning: In supervised learning the machine is provided with labelled data that is data tagged with correct answer. Then the machine is provided with new set of data, the supervised learning algorithm then use the training data to determine the correct answers.

For example, Suppose we have a packet of vegetables with different vegetables like tomato, potato, lady finger, etc in it. Now we have to train the machine for each kind of vegetable like round and red colored vegetable is tomato, green and elongated shaped vegetable with pointed ends is lady finger and so on. Then if we provide the machine with new photos of the vegetable , it will use the shape and color to determine what type of vegetable it is.

Application: Speech Recognition

In speech recognition we provide different words and sentences to the machine along with the name of the persons saying them. Then we give some unforeseen data to the machine, the machine will use the training dataset and with the help of which it can find the person whose voice was given to it. Supervised learning was used because the machine needs some correct data so that it can use that data to classify the voices.

Unsupervised Learning:

In unsupervised learning machine is provided with data but the data is not labelled with correct answer or labelled. The machine groups the data on the basis of its features by itself. The machine work on its own to discover the information.

For example, if we provide the machine with a dataset of different kinds of birds, then machine will classify each bird on the basis of their features. Like peacock has colourful, round feathers, toucan has colourful beak. So the birds with round and colourful feathers will be sent in peacock category and so on.

Application : Cancer Detection

The machine will be provided with data in which persons with different types of cancer tumors will be there. The unsupervised learning algorithm will classify the persons according to tumors condition whether it is dangerous or not.

Semi-supervised Learning:

In this type of learning the machine is provided with a very small correctly labelled data along with large unlabeled data. It falls between the previous to categories. To deal with the disadvantages of both supervised and unsupervised leaning, the concept of semi-supervised learning was brought into account.

Application: Speech Analysis

In supervised learning for speech recognition we would have to provide a large amount of labelled data for a single person. To overcome this disadvantage we use semi-supervised learning so that we can provide a small amount of labelled data and get the results for the unlabeled data.

If the answer helped, please upvote. For any query, comment.