I am trying to refine my knowledge of AI, but unsupervised learning is a bit of a stumbling block for me. I understand that it finds 'hidden' patterns in data, but if they are hidden, how does a user interpret the outcomes? It would be like someone categorising a deck of playing cards in some way, but the logic of that process is never known. How exactly is this helpful?

Onto my second question, which might help me understand the first question a little more clearly. What examples have there been in the real world of unsupervised learning being used, and what exactly did this neural net help solve?


2 Answers 2


A good example is face recognition on your phone,
or in face recognition systems in general.

The way they work is pass in infromation and tighten the channel width for information throughput, performing a compression of the input face features.

enter image description here

Then try to reconstruct the input on the other side by learning a decompression transformation.
If you're able to recreate the input features perfectly with less vectors representing that input, then you've learned a more ideal and abstract representation of the facial features.

The reason this is considered unsupervised is because the human is not trying to influence how it does this using their own reasoning. It's [face in --> compress --> decompress --> face out --> is face in the same as face out]. It's not "learning to find patterns". It's learning to create better representations.

A network that passes through, compresses and decompresses faces accurately is useless, but the intermediate compressed representation is VERY valuable. You'll hear about the compressed representation refered to as an embedding.

Facial recognition systems don't compare noisy input, but the pure representations that filter out noise and capture abstract representation. So well that the same face will end up with compressed representations that resemble that same face taken in slightly different angles, with different hair styles or colors. etc.

If a face embedding vector is close spatially to another face vector then it is similar in the abstract, not just circumstantial, sense.

You can now use spatial lookups on compressed face representations to do automated watchlist queries on camera feeds. If a camera face, when compressed, looks similar to a watchlist face, then send an alert.

Similarly you can compress the face on an ID, the see if it's similar enough to a face extracted from a selfie to do two factor facial/identity verification.


Unsupervised learning aims to find similarities and patterns in unlabeled data. It is used for clustering and association problems (you can google for an explanation).

One example can be a music recommendation system (e.g., Spotify): From the list of songs that you like, unsupervised learning can identify songs with similar characteristics that you may like as well. It can be based on the melody, tempo, and dynamicity of the song, etc.

Another example is online store recommendation systems (e.g., Amazon): Unsupervised learning can discover interesting relationships in large databases. For example, people who buy a table will most probably buy chairs and a sofa. People who buy milk usually buy bread and butter, and so on.

Another example is anomaly detection (fraud detection, health monitoring, defect detection). From a large amount of data, unsupervised learning can detect cases that differ from a norm (e.g., sudden burst or decrease in activity).

  • 1
    $\begingroup$ I'm not sure if recommender systems are the best examples of applications of unsupervised learning, unless you consider RL and bandits unsupervised learning, which some people do. Anomaly detection is a good example of a problem that can be solved by unsupervised learning techniques, but you can also use supervised techniques for anomaly detection. $\endgroup$
    – nbro
    Feb 1, 2023 at 23:12
  • $\begingroup$ Your answer could be improved with additional supporting information. Please edit to add further details, such as citations or documentation, so that others can confirm that your answer is correct. You can find more information on how to write good answers in the help center. $\endgroup$
    – Community Bot
    Feb 2, 2023 at 9:34
  • $\begingroup$ "It can be based on the melody, tempo, and dynamicity of the song, etc." More realistically, recommender systems are usually based on what other users like. If I like songs A and B, and most people who like songs A and B also like song C, then the recommender system is going to recommend song C to me, no matter its melody or tempo. $\endgroup$
    – Stef
    Apr 2, 2023 at 17:55

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .