By classical, I mean the current Machine learning Algorithms we have, according to the current status of Machine Learning field, some we have or might have not gained the in-depth aspects which will/might influence the enhancement of these classical Machine Learning algorithms.

On the other side, quantum machine learning will solve computational complexity (like space complexity, time complexity or communication complexity), so as to achieve AI goals or objectives.

Will quantum machine learning be achieved besides the computational power that is slowly advancing?

What is the difference (any idea on a global perspective) between classical machine learning and quantum machine learning?

Note that, typically, the semantic difference between the two terms would be subtle enough to be irrelevant today, but if we are interested in the growth nor improving the ML algorithms due to Big Data generation, then the distinction would be important.

  • 1
    $\begingroup$ What do you mean by "quantum machine learning will solve computational complexity"? Computational complexity is not a problem, but a sub-field of computation science. $\endgroup$ – nbro Aug 24 '19 at 13:20
  • 1
  • $\begingroup$ @nbro I copy that...,can we create a chatroom for this topic.So as we get it clear. $\endgroup$ – quintumnia Aug 25 '19 at 17:44
  • $\begingroup$ @FranckDernoncourt ,I've self-taught myself a hell of a lot too, I'm saying this from experience,until this out weigh my study knowledge.So i kindly request you to analyse this question inline with the answer we have so far! Though,there's a trillion localized real-world pieces of research information. $\endgroup$ – quintumnia Aug 25 '19 at 18:23
  • $\begingroup$ @quintumnia Maybe you can start by clarifying where you copied it from. $\endgroup$ – nbro Aug 25 '19 at 20:06

Classical machine learning algorithm is based on Bayesian statistics which can be calculated on Commercial off-the-shelf hardware (COTS). The typical workflow in creating a machine learning model is to use the Python programming language together with a GPU for determine the weights of a neural network and after a number crunching session the neural network is able to detect images. In contrast, quantum machine learning runs only on a quantum computer which is based on quantum gates.

It's not possible to train a quantum model on a COTS graphics card. The reason is, that the RAM of conventional computers can store only discrete information which is either 0 or 1. The requirement for storing qubit states, which are equal to a Bloch sphere representation [1], is a superconducting niobium processor. Because quantum computers are expensive, it's not possible for amateurs to test their machine learning models. Only in the future, the promising technology of quantum machine learning will become relevant for students and universities.

[1] Wikipedia Bloch sphere, https://en.wikipedia.org/wiki/Qubit#Bloch_sphere_representation

  • $\begingroup$ ML is not just based on Bayesian statistics. Stop giving answers or asking questions (aka, in your case, spamming). $\endgroup$ – nbro Aug 24 '19 at 13:16
  • $\begingroup$ ML is equal to handle large amount of data. This is what nvidia GPU hardware and tensorflow like libraries are about. The subject for dealing with raw data from great complexity is statistics. Quote from Wikipedia: “Statistics is the discipline that concerns the collection, organization, displaying, analysis, interpretation and presentation of data.” For reason of simplification it make sense to focus on the most successful part of statistics which is working with probability distribution. $\endgroup$ – Manuel Rodriguez Aug 24 '19 at 15:40
  • $\begingroup$ Your comment is quite misleading (like your answer). There is no such thing as a most successful part of statistics. Furthermore, I advise you to stop using the expression "X is equal to Y", which is highly misleading and imprecise. $\endgroup$ – nbro Aug 24 '19 at 15:58

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.