Well, a rudimentary look would indicate you just add 1 — and so our function is:y = 1x+1And there you go.
We found m=1 and b=1.
We just took some data (the table above) and created a function that described it.
In essence, that is what machine learning is.
We took our input x and made a prediction on what y will probably be for all examples.
The real fancy part is just how you teach a machine to learn what function best describes the data — but when you’re done what you’re left with is y=mx+b (on crack).
Once we have that function, we can also plot is on a graph as seen below.
MathIf you want a more comprehensive explanation of functions, here’s a site that looks like their web designer quit sometime in the early 2000s: https://www.
mathsisfun.
com/sets/function.
html.
It’s also called “math is fun” which is a potential red flag, but it’s intuitive and straightforward.
Humans can’t do the math machines canSo obviously y = 1x+1 is a bit of a ridiculous example.
The whole reason we have machine learning is because humans can’t look at millions of data points and come up with a complex function to describe the output.
Y’all watch Maniac?Again— this is all based on data.
You need enough data so that you can even learn such a function.
If we only have one example for x and y, then we would have no idea what the function is.
If we take the first example: x=1 and y=2.
Well then, our function could be y=2x……y=x+1……y=((x+1)*5–9)⁵ + 1……or any number of more ridiculous things.
If we don’t have enough data, the function we have our machine create is liable to have a ton of error when we try and use it on more data.
Additionally, in the real world data isn’t always so perfect.
See the example below where we can have a function that fits the data best, but the line will never pass through every point.
Unlike the old tables from math class, data collected from the real world is more unpredictable, and can never be described perfectly.
This is a basic example of a machine learning a function that best represents the dataFinally, the last thing humans can’t do is look at a bunch of variables.
It’s easy with just x and y, but what if there isn’t only one input variable?.What if y is affected by x¹, x²,…x¹⁰⁰.
Very quickly functions can become increasingly complex.
Machine learning in the real worldLet’s look at an example that isn’t hella abstract and hard to visualize because I’m sure everyone’s getting mad bored.
I work in pharma, so let’s use cancer as it’s an area of personal interest and exploration for me.
Say we have a dataset that has two variables on tumor size (radius and perimeter) and then a variable indicating whether or not the tumor is metastatic (potentially life-threatening).
This sounds like a way more difficult use case, but it’s a very similar concept!y now becomes diagnosis, and it can be 0 or 1.
0 is benign, and 1 is metastatic and dangerousx¹ now becomes radiusx² now becomes perimeterWe have an m for each x, I’ll call them somethingb stays the sameSo how does our linear equation look now?.Not that different!diagnosis = (something1*radius) + (something2*perimeter) + bNow like I said, we’re getting out of the realm of human capability.
So instead of looking at the data and trying to find out what something we have to multiply our variables by to get an accurate estimate of diagnosis, we have machines do it for us.
And that is machine learning!Of course, this data isn’t perfect and therefore our model won’t be either.
While it won’t be right 100% of the time (just like a human), we simply need the machine to come up with the best possible function it can, that’s right most of the time.
There are tons of other considerations when having machines learn to detect cancer, but hopefully this real world example shows the considerations of what variables you have your machine look at, and how important the data you learn your function from is.
I mean yeah there’s a lot more…Remember this memeOkay so this was definitely a scratch on the surface of that incredible mathematics and computer science that goes into machine learning.
But still, the concept is the same.
No matter how wildly impressive or ridiculous, all machine learning stems from functions learned to best describe data.
The real question is — is that how you perceive human consciousness…?.lol.
.. More details