Deep learning | Machine learning | AI -Clearing the confusion.
If you’re not confused, you’re not paying attention. — Tom peters
Deep learning | Machine learning | Artificial Intelligence
These are the terms which have confused a lot of people across & if you too are one among them let us together resolve it!
Being a programmer or a geek its important to use these terms wisely because THEY ARE NOT SAME BRO!
Artificial Intelligence : A wider umbrella.
AI is the wider umbrella under which ML or DL (Deep learning) come. Keeping it simple, We can say that all 3 are related with other in some or the other sense.
|Understanding the difference :
AI : Makes the machine learn through experience. Machine adjusts their responses based on new input and thereby performing human like task. They can be trained to accomplish specific task by processing large amount of data and recognizing pattern in them.
Example: We can think of AI as constructing a huge building, the first building took years of years to finish. So most of the labors never saw the final outcome. Those working on it took pride in their craft, building bricks and chiseling stones, That were going to be placed in that giant structure.
So as AI researchers we should think of ourselves as humble brick makers whose job is to study how to build components, learning the procedure/algorithm etc. That someday, somewhere,someone will integrate those blocks into the intelligent systems. Like, bricks and cement components result into huge buildings.
Machine Learning: Make machines learn
After clearing the understanding about AI , Its time to get confused and dive deep down to machine learning.
The first question comes to mind after understanding AI is “What made ML come into existence when we already had AI??” So, to answer this question- its important to discuss the problem faced by people with AI.
i. How to efficiently train large & complex models?
ii. How to train more robust versions of AI systems?
iii. Researches faced the problem to design the operation model of the brain?
These issues had the higher influence which let machine learning came into existence.
Machine learning have methods and model whereas AI had symbolic approaches. Machine learning enables the computer to take data driven decisions and carry out a certain task. Which trains itself on the certain models and improves when exposed to the new data.
Example: Suppose you go to the market to buy mangoes and there are 1000’s of mangoes there. So, it becomes difficult for you to choose which one is the best. So let’s say you choose a mango and concluded that the bigger and bright yellow mangoes are sweet, While the smaller and the pale yellow mangoes are sweet only half of the time. But some other day you go to market to purchase oranges but now you can’t use the logic of mangoes with them so, It is worthless at this point of time and so again you have to learn everything for oranges as well. But this is not as difficult as we think. But what if we have to program this logic? The problem is every time you make new conclusion you have to modify the rules manually which becomes difficult to read and understand and takes lot of efforts, research & time to code.
This is where now Machine learning comes into picture, Which allows the machine to learn from example and experience and that to without explicitly programmed. So instead of you writing the code, what you do is you feed the data to the generic algorithm/machine.
Deep Learning : Dive Deep into the ocean of learning.
Deep learning is not new, but nowadays it’s hype is increasing and getting more attention. You can consider deep learning model as a rocket engine and its fuel is huge amount of data that we feed to these algorithms. This field is particular kind of machine learning that is inspired by the functionality of our brain cells called neurons which led to the concept of artificial neural network.
It takes data connection between all the artificial neurons and adjust them according to the data pattern. More neurons are added at the size of the data is large it automatically features learning multiple levels of abstraction. And thereby allowing a system to learn complex function mapping without depending on any specific algorithm.
Example: Lets say we have a machine which recognizes whether the given animal image is cat or dog?
Now suppose if were asked to resolve this question with the help of ML approach. The first thing we would have done is that we would have defined the features of the given animal such as, check that it has whiskers or not? or check whether the animal has pointed ears or not? or whether the tail is curved or straight? In short we would have defined features and let the machine identify which features are more important and classifying it into particular animal.
Now when it comes to deep learning it takes this into one step ahead. It identifies automatically which features are most important for classification. Comparing to machine learning where we have to manually give out the features.
Till now we can conclude that AI is the wider umbrella under which ML and Deep learning reside.
The 2nd conclusion we can make is “Deep learning IS Machine learning” with some advanced versions and features i.e it is the next evolution of machine learning.
|Summary:
Machine learning uses algorithms to parse data, learn from that data, and make informed decisions based on what it has learned.
Deep learning structures algorithms in layers to create an artificial “neural networks ” that can learn and make intelligent decisions on its own
Deep learning is sub field of machine learning. while both fall under he broad category of artificial intelligence, deep learning is usually whats behind the most human-like artificial intelligence.
