Nathaniel3W
Member
I was contacted by a government contract company. If you're a government employee, and you give a contractor your name one time, they'll contact you for the rest of your life. They had just landed a big contract and were looking for more people to fill it. They wanted my latest CV and salary expectations to see if they could place me. I told them my desired salary and I honestly thought they wouldn't respond... But then they did.
So now I'm waiting for interview number 2 and I'm brushing up on some things so I don't sound like a complete idiot.
My experience with AI and machine learning is basically understanding the concepts, and then importing a library and using it. I decided I really wanted to learn the details and I started writing a neural net from scratch. Right now it has a lot of loops that I'll eventually want to rewrite for parallelization. And there are a lot of places where I wrote code, and the book's code is a lot cleaner because the author points out "But as you can see, this is the same thing as a dot product, and we can replace this whole part..." and he also writes code very pythonically, whereas I just don't think like that and I have to see the extra steps written out. But the biggest challenge right now is this one very important backpropagation function/formula that the writers of these different books keep putting off. They keep explaining it conceptually, and calling a function to do this thing, and saying, "Trust me. This is what it's doing." And it's the most important part, and it's the one part of my neural net that I can't write from scratch yet.
So now I'm waiting for interview number 2 and I'm brushing up on some things so I don't sound like a complete idiot.
My experience with AI and machine learning is basically understanding the concepts, and then importing a library and using it. I decided I really wanted to learn the details and I started writing a neural net from scratch. Right now it has a lot of loops that I'll eventually want to rewrite for parallelization. And there are a lot of places where I wrote code, and the book's code is a lot cleaner because the author points out "But as you can see, this is the same thing as a dot product, and we can replace this whole part..." and he also writes code very pythonically, whereas I just don't think like that and I have to see the extra steps written out. But the biggest challenge right now is this one very important backpropagation function/formula that the writers of these different books keep putting off. They keep explaining it conceptually, and calling a function to do this thing, and saying, "Trust me. This is what it's doing." And it's the most important part, and it's the one part of my neural net that I can't write from scratch yet.