What is machine learning? The future scope of machine learning and how to get started with Machine learning.
Well, machine learning is an old term but it has gained its popularity recently. As we all know that in this digital world we are producing a lot amount of data which is in trillion of GB.
WHAT IS MACHINE LEARNING ?
Machine learning is a subset of artificial intelligence in which we train machines to learn things with the help of various complex algorithms.
As we all know human learns from their experience, in machine learning we use data to train machines or in short Machine learning is a type of artificial intelligence that provides the computer with the ability to learn without being explicitly programmed.
Steps of machine learning :
1. Collecting data = The first thing we do in machine learning is collecting the data. Without data, we will not be able to train our ml model.
2. Cleaning the data = The second most important thing we do after collecting data in machine learning is to clean it. For our ml model to minimize the error we have to give it clean data.
3. Analysis of data or labeling data = When we are done with collecting and cleaning the data we have to label or analyze it before feeding it to algorithms. For example, we have a data set of company sales. Then we have to label sales according to months.
4. Train the algorithm = In machine learning after labeling data, we had to train our algorithm that can be linear regression, k-means algorithm, etc. by accurate data so that our algorithm can learn from it, so what we do is we give 80 % of our trained data (Trained data = the data that has been already processed ) to the algorithm to train it.
5. Test the Algorithm = After training the algorithm we had to test whether it works accurately or not. So for this, we take the remaining 20% of our trained data and feeds it to the algorithm, and compare its outcome to the actual result. If the outcome has too many errors then it means our model is not efficient we have to do several modifications to it.
6. Use it = After training and testing we can apply our model to various data.
Types of machine learning : Basically machine learning is of 3 types.
1. Supervised learning machine learning = In supervised machine learning first, we have to label data before feeding it to the algorithm.
2. Unsupervised machine learning = In unsupervised machine learning, we use labeled data to feed to the algorithm.
3. Reinforcement machine learning = In reinforcement machine learning our machine learning is been programmed that kind of that it can label the data itself and able to understand it.
Artificial intelligence vs machine learning vs deep learning.
So let's start with deep learning. Well, deep learning is a subset of machine learning. In deep learning, we use a deep neural network to train machines. A neural network is based on the human brain neurons. No one knows what happens in a neural network. The problem with deep learning is that it takes a huge amount of data to train itself and it can take about weeks to months to be fully trained but the plus point with deep learning is that provides high accuracy as compared to machine learning.
Okay let's take this question ai vs ml. So in machine learning data, we train algorithms for a particular work. For example, if your trained machine only knows about playing cricket then it can only play cricket but if you put football in front of it and tells it to play then your model will fail absolutely because it is only trained to play cricket.
Whereas if there was an Ai then it tries to play football all this means that artificial intelligence is capable to think itself to do things without the need of any human interference.
This is all about the machine learning.
0 Comments