Hello All , It’s been long time i haven’t written anything in my blog. I have just now started learning “Machine Learning” . I will try to write a post about my learning as much as possible .
Why I choose Machine Learning?
Machine Learning came some long time ago , why i choose to learn now? The Information technology world is changing every single day . Yes “Change is the only thing constant” in this technology world .
Even though i have started my career as software developer , it was in my long term list to learn machine learning , i didn’t pay much attention to it . Now i have decided to learn out of my passion and interest for it .
How am i learning Machine Learning?
I am learning Machine Learning from videos and text books . I haven’t taken any course for that . I have started watching Andrew N videos and few texts from others to learn .
What is Machine Learning?
It’s very difficult to explain in words , i will try to explain in my own words
Machine Learning defines ability to teach/train machine(system) to predict certain things based on the rules given to them and also improves efficiency based on the experience .
Best real time example is “Playing chess against computer AI” . In AI it has rules( how pieces can move , how many steps can move) , with that it predicts and makes a move for a human move. In short there is no hard coded way if this move then this move , It calculates all possible moves against a move and makes the best move . How this is been done? System is loaded with tactics and statistics for a game . It improves from it’s experience and becomes better at it .
Let’s get to the introduction of what we are going to see in this .
We can divide the Machine Learning into four important Parts
- Supervised Learning
- Unsupervised Learning
- Deep Learning
- Reinforcement Learning
We will see each one with small introduction to it , in later posts we will explain each one with detailed examples .
Supervised Learning
In a given data (known labels) find meaningful relationship between them and predict/estimate the output. We can split Supervised into two categories
- Regression
- Classification
Regression:
For example: Predicting House Price for a given square feet .
Lets assume this data set 1000sqft -> 4Lakhs, 2000 -> 7.5 lakhs , 5000sqft -> 18lakhs .
When asked to find the price for 700Square feet , we can easily draw and map the price from given data set . This is called regression , because the value we are trying to predict is continuous.
Classification :
This another type of supervised algorithms . Best example for this type is predicting /identifying blood group of humans .
Here Y axis value is constant ( O+, O-, A+, A-, B+, B-, AB+, AB-) . But we can have discrete number of x values (Human blood samples) . Here Y can take discrete values of X hence it’s called classification Supervised Learning .
Unsupervised Learning:
Unsupervised learning defines , analyses unlabeled data and groups/clusters them .
for example:
In above image , as you can see given unlabeled data , unsupervised machine learning algorithms analyzes and finds pattern and groups/clusters them together . This is called Unsupervised learning where unknown data given .
We will see about other two types also in detail once we finish these types of learning in detail in upcoming posts.
That’s it . Thanks for reading . All kind of feedback are welcome. Happy coding !