Machine Learning is a part of Artificial Intelligence. And AI is a very broad technology and there are basically two main sub-parts of it:
- Machine Learning
- Deep Learning
So we can say that if you want to learn and understand AI then you should have the clear-cut knowledge of ML and Deep Learning. Because AI is internally dependent on these two main sub-parts.
What is Machine Learning
ML is an application of Artificial Intelligence that provides computers the ability to automatically learn and improve from experience without explicitly needing to program.
In ML, we clearly focused on how to create such a machine that can learn something by itself by observing data (such as images, locations, objects, facts, etc.).
ML Training Process
The Machine Learning process can be done in four stages:
- Take a huge amount of data about that object for which you want to apply the ML Model.
- Define that object’s characteristics, features, functions, etc, of that object.
- Choose which machine learning model you want to use to get filtered data.
- Pass that whole data (such as images, characteristics, functions, and features) through that machine learning model and gain the final answer.
Let’s understand the above four stages with the help of one example. Here if I want my computer to identify whether the picture is of a cat or not. Then let’s take a look at how this thing will work:
- I will take many images in which there are many cats in different positions, the different breeds, in different colors, in different situations, etc.
- Then I will define the characteristics of a cat. Like how basically it looks, its eye shape, its average height, and what is its functions and all of those things.
- Then I will choose the machine learning model which I want to apply to data to gain results.
- In these final steps, I will put that data on that machine learning model to get the final answer.
So now after applying all those steps, my computer is ready to identify whether it is an image of a cat or not. So now you may have understood how complex a process is. This is just applicable to a cat, imagine how many objects, situations, and things are there in the real world. So we can say that we have a lot of work to do.
Google Lens, Face unlock, and AI-based Cameras are one of the latest examples of the application of ML.
Future Path of ML
Today we have a huge collection of data sets and ML libraries (like NumPy, TensorFlow, and PyTorch). We use them for creating much more powerful ML models.
But still, ML needs much more amount of human interaction to get its job done. The main goal of ML is to create a system that can learn by itself. So we need to find out the way to automate the above four stages of ML!! Where computers can automatically get the data, understand its characteristics and filtered it, choose the right ML model and apply that and understand something on their own and that’s the future friends…
That’s it for today guys. Thanks for reading. Stay tuned for more amazing content.