101 : Building Blocks

DataVerse
DataVerse
Sep 1, 2018 · 3 min read

What are the most used words in the AI spectrum ? Well , there are many . But some of the most used words are also, unfortunately , the most abused words . There is so much media hype around the entire buzz AI has generated . It has definitely shown a lot of promise with the significant improvement in a lot of factors in the past few years . No one can deny the fact that AI is here to stay. Now , on one hand that might be a super awesome thing but on the other hand it is equally important to understand the nitty-gritty details about the entire subject.

Take Away from this page : Understanding the most widely used words , what these are and what these aren’t .


Artificial Intelligence : AI , as it is so fondly called , is actually the intelligence that machines demonstrate . Hence , AI research is a branch of science that sits at the intersection of many theoretical disciplines like Computer Science and Mathematics and aims at providing machines the ability to take decisions based on certain parameters. Artificially intelligent systems can rely on many techniques to be able to take decisions including that of Rule Based Systems . In fact , rule based systems have been an integral part of the entire AI journey which unfortunately is not a very well popular fact.

Machine Learning : AI systems don’t just rely on rules . That’s where the concept of ML comes into picture. Machine Learning gives computers the ability to learn patterns without a programmer having to explicitly program all the nuances . That is different from a rule based programming approach and the below graphic explains it so well.

Source : https://github.com/mahmoudparsian/data-algorithms-book/tree/master/src/main/java/org/dataalgorithms/machinelearning

In a traditional programming paradigm , the computer accepts an Input and a program which enables it to generate an output . While in the ML world , the computer takes Input and Output and it learns what program maps the input to the output.

Deep Learning : Deep Learning is a subset of Machine Learning that is based on learning data representations as opposed to a task specific learning. In traditional ML models , we need to tell models the features to work on , while in Deep Learning , the features are auto identified and thereafter used . Deep Learning has proved to be unprecedented in showing improvement especially working with unstructured data like image and text . It has bagged State Of the Art results on tasks like Image Classification , Object Detection and Question Answering to name a few.

AI , ML and DL are words that are often spoken as completely isolated terms. While in practice , these are just of same family. The below graphic shows how three are related to each other

Source : https://towardsdatascience.com/cousins-of-artificial-intelligence-dda4edc27b55

AI is the super set and both Machine Learning and Deep Learning are subsets . Further , Deep Learning is a subset of Machine Learning. Understanding this will help you not say things like ‘You just used a machine learning model , when are you gonna use AI’.

DataVerse

Written by

DataVerse

Space for discussing algorithms that make machines intelligent

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade