标签：snippet else nts rmi break each currently variables imp
What exactly is Machine Learning?
You must be thinking that wait this doesn’t add up, you were told difficult definitions with heavy technical words. We will break them down one by one. Just like you attend classes and learn concepts the same is done with machines, the data is fed to the system and the machine is ‘trained’ on this set. It is also referred to as the ‘training dataset’. Now just like you sit for exams and evaluation similarly the machine is also ‘tested’ on the dataset which is referred to as the ‘test dataset’.
If our answers in the test are correct that means that the accuracy between us attending the classes and grasping knowledge versus us using those taught concepts correctly and finding an answer is very good. Similarly, our model is good if we are able to provide accurate results after training it.
So picture this, you attend a Math class and learn how to solve problems. You then sit for its exam and solve different problems than the ones you did in class but use the same concepts. This is Machine Learning except that the ‘machine’ is you.
Once we get our results we know where we need to improve and where we lost marks, exactly like that we test the ‘accuracy’ of our model on the machine and keep improving it to provide better results.
Another analogy and my favorite one for better understanding this could be that think of it as you’re in a room and you don’t know where the exit gate is. So what do you do? You start observing other people and where they are going to find that exit. So I’m actually looking at other people for my course of action, the same goes in machine learning wherein we feed these pre-existing paths to exit for the computer and train it and then expect it to find out its own exit path. Thus, Machine Learning Course is now present in abundance to teach the humankind, the beautiful prospects of how does the functionality of Machine Learning actually works.
What does Machine Learning have in store for the human mankind?
So broadly there are two categories ‘Supervised Machine Learning’ and ‘Unsupervised Machine Learning’.
To understand these, we first need to learn the terminologies i.e. Labels and Features.
Features are all the variables through which you get trained in your Math Class, these could range from your concentration to the teacher’s dedication and the list goes on. The same applies in Machine Learning, features will be all those variables that help you to better train your model.
Moving on to Labels, these are variables that has the output. So in the analogy of exams you can think of them as your final grade/marks. Here is a small python code snippet to get things a bit interesting.
if(labels==exist):
print(“Supervised Machine Learning!”)
else:
print(“Unsupervised Machine Learning!”)
Supervised Machine Learning
This is made up of ‘Regression’ and ‘Classification’. Think of regression as prediction and classification as categorization.
For now, just understand how we are putting the same kind of data into the same category in case of classification but trying to predict from this data in case of regression.
We will be going through each of these separately. This is the reason why the Machine Learning Tutorial has so much common people searching for it.
Unsupervised Machine Learning
This is made up of ‘Clustering’ and ‘Association’. Think of clustering as forming different clusters of similar data in it and Association as the relationship between the different data.
Regression
A case of a simple linear regression model in which we just have one independent variable and one dependent variable.
If we increase the number of our independent variables, then it is called multiple linear regression depicted by the following equation:
Other regression algorithms:
Applications:
Classification
The machine learns from the pre-existing data and classifies new data in the existing categories
Classification algorithms:
Applications:
Clustering
Group similar kind of data together to understand their classes that you are not currently aware of and find out internal pattern and other useful information.
Clustering Algorithms:
Applications:
Association
Given a set of transactions, we can find rules that will predict the occurrence of an item based on the occurrences of other items in the transaction.
Association Algorithms:
Applications:
Machine Learning Changing the world in fields
What Machine Learning Have In Store For Normal People?
标签：snippet else nts rmi break each currently variables imp
原文地址：https://www.cnblogs.com/data-science-course/p/12002658.html
