Machine learning is super hard and frustrating. I have no idea where to start 😓

Start with udacity's ud120 & then go for Andrew Ng's course on Coursera. This should be enough to get started.

And always remember, machine learning is fucking hard, no matter what you do.

@vish + I'm in 8th grade. Sure i have A in math but i still don't understand a single thing.

It will probably start with an if-statement and then another, and one more, and so in 😉

@krister-alm xD trying to learn tensorflow :)

@Dacexi you are going to have to learn some higher level math first

8th grade?First Learn math and calculus kid

@bdhobare @nickhh yeah, i do realise that.

You know "A" and "B" and you want the machine to get to "C" and eventually "D".

You should just let the computer learn it instead. *ba-dum-ptss*

Codewars: binary genetic algorithm. Start from the basics

Machine learning is a fascinating but huge can of worms. To understand what you do you should probably start with linguistics (Yes, no joke, formal understanding of semantics is invaluable), Statistics, Calculus and ideally quantum mechanics (You'll learn many approaches to distill a factual result from sets of probabilities). I'd start with genetic algorithms since they're quite intuitive and bayesian beliefs networks, they're incredibly powerful and relatively easy to understand (meaning the probability of coming across complex Calculus is avoidable).

@Godisalie thank you for mentioning quantum! I've tried bringing up the similarities to colleagues before, but got so dissuaded with the overly negative responses that I gave up on studying machine learning and even quantum itself (was a computational condensed matter grad student) for almost two years, just before it all got big.

@USER-NOT-FOUND I've studied Chemistry with focus on computational physical chemistry for 5 semesters and the tool box you need for quantum mechanics is basically the same you need for machine learning. Both measures probabilities and need to project it into a state, I hardly see how anyone familiar with the concepts could deny the similarities. Sure, the sets of measurements are vastly different but the process of evaluation comes down to the same methods.

Start with 'why'.

Using a tool that's more complicated than your problem is silly and a waste of time. Sure, build a toy model if you want, but don't go bragging about it.

It doesn't matter how hard you worked, it matters what problem you've solved. If you're in grade school and are using artificial learning, it's probably because you don't really understand the problem you're trying to solve.

Start with a problem and look for a solution, not the other way around.

.

@samxxx make it two

.

@Dacexi I'm 16 and honestly it took a few hours through github and computerphile YouTube videos to understand how to implement it in code. It doesn't take a genius, it takes a creative one!

@digital-dina yeah I've started to understand it a bit now. I've managed to make an image classifier and an LSTM that generates Shakespeare-like text :)

@Dacexi That's actually impressive. I've only managed semi-plausible YouTube comments lol