Deep, Machine Learning and AI. What are your differences?

Asked

Viewed 176 times

8

About Deep Learning, Machine Learning and Artificial Intelligence, what are your main differences?

  • What are your principles?
  • They are all who can learn and solve problems of their own nature?
  • How to distinguish one from the other?
terminology artificial-intelligence machine-learning
  • 1

    I don’t understand exactly what you mean by "solving problems of your own nature"

  • @Arthurferraz meant about the machine identifying problems and solving them without a totally explicit programming there.

  • 1

    ML techniques can solve problems without being explicitly programmed to do so. I just don’t understand exactly what it would be like to identify problems.

  • @Arthurferraz machines that identify situations that they themselves can handle.

  • DL is a specific ML model, and ML is made as one of the possible techniques to enhance the fitness function of an intelligent agent, aimed at getting a good weak AI

  • If it’s in code it’s Machine Learning, if it’s in Powerpoint it’s Artificial Intelligence :P :D

Show 1 more comment

1 answer

9


The basic difference between the areas mentioned is in granularity. Deep Learning is a sub-area of Machine Learning, which in turn is a sub-area of Artificial Intelligence.

That is, all Deep Learning technique is also Machine Learning and, therefore, Artificial Intelligence. Not every AI algorithm is ML anymore, and not every ML algorithm is Deep Learning.

Artificial Intelligence

Defining Artificial Intelligence is quite complicated by the fact that the term intelligence is not well defined. But briefly it would be the area that studies algorithms that seek to mimic human intelligent behavior. So any algorithm that claims to be intelligent in some aspect would be considered Artificial Intelligence.

I can cite optimization algorithms such as 'Finding the best possible route through N points' as an AI problem that is solved without using machine learning.

Machine Learning

The confusion between AI and Machine Learning (ML) nowadays is quite common because most of the successes and advances of AI are linked to ML techniques, however this area is corresponds only to a part of the algorithms.

ML is commonly defined as set of AI algorithms that are able to perform a procedure without being explicitly programmed for it.

We can cite as techniques ML Decision Trees, Bayesian Networks and even the infamous Page Rank (Google ranking algorithm). However, all these cited nay are deep learning techniques (DL).

Deep Learning

One of the most successful techniques of ML has been neural networks. Due to the increased processing capacity it was possible to build deeper and deeper networks, giving rise to the term Deep Learning.

Deep Learning would then be the area within Machine Learning that specifically studies Deep Neural Networks.

The most popular DL techniques I can mention are Convolutional Neural Networks (very useful for solving problems with images) and LSTM (useful for solving problems where there is order dependency, such as text generation).

Reference: https://propus.science/diferenciacoes-entre-inteligencia-artificial-machine-learning-e-deep-learning/)

This Thread has a great explanation on the subject: What is the definition of Machine Learning (Machine Learning)?

Any doubt may ask that I enrich the answer.

  • Good the clarity and explanation of the three items, but did not understand relationship. However much in the explanation, it has been clarified where each acts in the other, does not make clear what differs from each other, nor its principles.

  • I edited the answer to make it clearer. But in general the important thing is to understand that Deep Learning is a sub-area of Machine Learning which, in turn, is a sub-area of Artificial Intelligence.

Browser other questions tagged terminology artificial-intelligence machine-learning

You are not signed in. Login or sign up in order to post.