Machine Learning : Supervised vs Unsupervised Learning

Artificial Intelligence is a branch of computer science that enables a machine, to perform a task similar to that of humans or human intelligence. Machine learning is one of the sub-fields associated with artificial intelligence. Machine Learning is an algorithm that learns from the experience without being explicitly programmed.  Deep Learning is a subfield associated with Machine Learning that leads to creating a neural network that emphasizes working similar to neurons in the human brain.

According to Computer Scientist and Machine Learning Pioneer Tom M Mitchell, "Machine learning is the study of computer algorithms that allow a computer program to automatically improve through experience".

Some of the real-world application associated with Machine Learning:

1. Image Recognition

Image Recognition can detect various objects within the photographs or images. It can identify the object taking into consideration various factors such as the intensity of pixels in black and white or colored images.

Example:

Identify the image composed of a dog or cat or any other animal.

Identify the cancerous cells in the pathological scans.

Recognize the Handwritten Numbers.

2. Speech Recognition

Machine Learning can translate or convert the speech into text and vice versa. Modern-day smartphones such as Apple iPhones composed of virtual assistance as SIRI which recognize the voice of the owner of the phone. Voice assistance includes CORTONA, AMAZON ALEXA.

Example

Voice Search

Voice Dialing

3. Predictive Analytics

Machine learning can classify the data into various groups it can be binary (classify into 2 groups) or multiclass (more than 2 groups).

Example:

Classify the news as fake or legitimate

Classify the email as spam or legitimate

Classify the tweet as abusive or non-abusive.

Supervised Learning

Machine learning is further divided into 2 categories termed supervised learning and unsupervised learning.  Supervised learning is the most popular paradigm. Supervised learning is teaching with question and answer

Example

Given a data (question) and label (answer). Machine tries to find out the various patterns within the question and how the answer is associated with it and using thousands and thousands of examples before reaching mastery to answer the question from the same domain.

In the above given example, the machine learning model(algorithm/estimator) trying to predict whether the image is composed of banana or apple. After a good amount of data (data composed of an image of fruit along with its label as apple or banana) being processed by the algorithm such as shape, color, texture, size of root on the top, the model is said to be trained. After training, the model is proficient enough to predict the given image composed of an apple or banana.  This is a simple example of supervised learning. Some of the well-known supervised machine learning algorithms are:

Logistic Regression

Decision Tree

Random Forest

Support Vector Machine

Naive Bayes

Gradient Boosting

Unsupervised Learning

Unsupervised Learning is another type of machine learning and the exact opposite of that of supervised learning. In unsupervised learning, the algorithm is given with data but not labels that only question no answers. Our algorithm is fed a lot of data and given tools to understand the properties of the data. After processing a sufficient amount of data, the algorithm starts forming the groups and differentiating the data into various groups, more technically termed as clusters.

Example

 

In the above given example, the machine learning estimator has given a huge load of data i.e(only question and no answers) or no labels. The machine learning algorithm starts processing those huge chunks of data, the algorithm is intelligent and powerful enough after certain processing, they try to group similar items into blocks termed as the cluster. This process is termed clustering. Some of the well-known unsupervised machine learning algorithms are:

K-means Clustering

DBSCAN

Hierarchical Clustering

References:

https://medium.com/towards-artificial-intelligence/what-is-machine-learning-ml-b58162f97ec7

https://www.technologyreview.com/2018/11/17/103781/what-is-machine-learning-we-drew-you-another-flowchart/

https://builtin.com/artificial-intelligence

https://www.salesforce.com/eu/blog/2020/06/real-world-examples-of-machine-learning.html

https://neurospace.io/blog/2020/08/what-is-supervised-learning/

https://data-flair.training/blogs/types-of-machine-learning-algorithms/

https://www.intel.com/content/www/us/en/artificial-intelligence/posts/difference-between-ai-machine-learning-deep-learning.html