You watched Terminator, right? At least one of the movies? If you havn't, go watch it.
What is Machine Learning?
To answer that, let's start with how we, the human beings, the mighty race, learn anything. If I teach you about how to make additions of any two numbers, you would know how to do after so many practices. So showing you "1 + 1 = 2", and "2 + 2 = 4", etc. You get the idea of how to do additions.Then when I give you a random 2 numbers, I expect you to give me a sum of these two numbers in return. Let's say "144" and "956". You would give me "1100" as the correct answer. If you get anything else than 1100, please retake your addition tests here.
The above scenario is what we are trying to do with machines. Teach them how to make perfect additions or perfect subtractions or just help us create a formula that gives us the desired outcome.
Why would we do Machine Learning?
OK, so you already know how to do additions and sum of whatever data has been thrown at you. Perfect! Now imagine giving you more than just 10 different numbers. Easy, right? How about 1,000,000 different numbers, where 500,000 adds the other 500,000. Not so easy, right? That's where we need the machine to do it for us.Let's try a much more complex example than just having a machine do basic addition. How about making the machine learn how to make a decision for us? Imagine a bin filled with red and blue marbles. We don't know how many red marbles and how many blue marbles there are in a bin. So this is totally random. Each time when we take a marble outside of that bin, we can immediately tell, that in our hand is what colored marble it is. Can we teach a machine to identify the different colored marbles there are in a bin and give us a total count of how many there are in a bin? Yes, it can. How does it do it? We don't know and we will never know. That is a target function that the machine will have to come up for itself with the given data input. This is just a general example of having a machine identify an object and give us an output. You could also think of this as a vending machine where it identifies what coins were inserted and give us a value output. Or a money counting machine, in a sense.
Examples of General Application of Machine Learning
Some great examples of Machine Learning is to predict the weather. By giving previous weather data, we can predict what would be the weather forecast for tomorrow and the following week and so on. There are many more examples but I leave that to your imagination and exploration.Can I get a job from mastering Machine Learning?
YES! We call these people, Data Scientists. They usually come from different specialization fields. Examples are:- Licensed Surveyors - They take all of the information and factors of an area around and in between a piece of land. Put it in a machine and it determines the value of the area and the projected value of that land area.
- Economists - They take all of the information and factors of the population, the quality of the workforce, the GDP, etc. Put that into a machine and it can forecast the rise and fall of the nation's currency.
- Business people or Bankers - Do sales forecast for the next quarter of the year, where each month, they have tracked how many sales made from January to April, and from there, they can predict how much sales is to be expected in the next quarter, or for the next year.
- Doctors - To determine the effectiveness of their new vaccine, they would need to run tests on different factors and information provided in a contained environment. Many mutations, different genomes, more generations of the vaccines developed, and lo and behold, a vaccine strong enough to cure the disease. Cancer is still unpredictable at this point since data is inconsistent.
Interested in becoming a Data Scientist? You can start here: