Difference between supervised, unsupervised and semi supervised learning with examples and applications
SUPERVISED
If
there is a training database and you have to identify the classes within the
training database, then it comes under supervised learning.
Examples
1. Check if a particular
image is of a face or of a car.
2. Check if an animal is
mammal or reptile
3. Predicting the
weather.
4. Bio-metric
attendance.
5. Given information
about the bungalow, predict the price.
6. Netflix – Given a
user and movie, predict the rating of the movie.
7. Finding images which
contains human faces- given a database for training
8. Identifying number
plate of a car, given a database of pictures and numbers.
9. Email is spam or not.
10. Identify the patient
is ill or not.
11. Predict stock market
price.
12. Predicting which
batch of food will taste bad based on factory temperature, humidity and
previous taste tester data.
13. Identify if a news is
related to politics, sports etc.
UNSUPERVISED
If
there are no classes and you have to group something without having a training
database, then unsupervised learning is used.
Examples
1. Grouping the pixels
by color in an image.
2. Given a news report,
group similar news together.
3. Find fraudulent
credit card transactions.
4. Find a yellow circle
in a video without having any samples of the video.
5. Find the most
effective hashtag for more likes on Instagram.
SEMI SUPERVISED
If
there is a training database and you have to identify different classes other
than from the training database, then it comes under semi supervised learning.
Examples
1. Given a training data
set of Ships, Identify tank in a picture.
2. Speech analysis.
3. Web content
classification.
4. Find cars, scooters,
and bicycles in an image given the training data set with only bicycles and
cars.
Usually classification and regression are considered as supervised
learning.
Clustering and association are unsupervised learning.
Also learn the Difference between Classification, Clustering and Regression
with examples and applications.