Supervised Learning vs Unsupervised Learning?

What is Supervised vs Unsupervised Learning?

Supervised learning is the machine learning task of learning a function that maps an input to an output based on example input-output pairs. A supervised learning algorithm analyzes the training data and produces an inferred function, which can be used for mapping next examples. An optimal scenario will allow for the algorithm to correctly determine the class labels for unseen instances. This requires the learning algorithm to generalize from the training data to unseen situations in a "reasonable" way. 

Unsupervised learning is a term used for Hebbian learning, associated to learning without a teacher, also known as self-organization and a method of modelling the probability density of inputs. A central application of unsupervised learning is in the field of density estimation in statistics, though unsupervised learning encompasses many other domains involving summarizing and explaining data features. It could be contrasted with supervised learning by saying that whereas supervised learning intends to infer an a priori probability distribution. 

What are the common algorithms under by those two categories?

Supervised learning algorithms:

Support vector machines Linear regression Logistic regression Naive bayes Linear discriminant analysis Decision trees K-nearest neighbor algorithm Neural networks

Unsupervised learning algorithms:

Clustering (hierarchical clustering, k-means, mixture models etc) Anomaly detection (Local outlier factor) Neural networks (Hebbian learning, Autoencoders, Deep belief nets etc)

More importantly, what we have to call it out is that the invention and development of unsupervised learning mostly depend on the advancement of academic evolution and limitation breakthrough of technologies, including the computing power, mathematical development etc. 

What are the biggest differences?

Compared to supervised learning where training data is labeled with the appropriate classifications, models using unsupervised learning must learn relationships between elements in a data set and classify the raw data without "help". 

This hunt for relationships can take many different algorithmic forms, but all models have the same goal of mimicking human logic by searching for indirect hidden structures, patterns or features to analyze new data. 

To understand it more psychologically, we could use an example as below to give you a glimpse of the differences between those two:

Supervised : "Here is what a cat looks like. Here is what a dog looks like. Now go classify some animals into those two categories. "

Unsupervised : "Here are some pictures of pets. Look for interesting classification structures in there and report back to me. "

 

 

 

 

 

First created & edit: AI Analytics

Data source: Wikipedia, Data toward science, Reddit