4 ms·
This makes me think I should upload my random computational intelligence assignments and get sweet, delicious karma. Edit: OK, I was a bit snarky, but I'm seri
by jacques_chester 14y ago
This makes me think I should upload my random computational intelligence assignments and get sweet, delicious karma.
Edit: OK, I was a bit snarky, but I'm serious.
I see these things on HN all the time. These are, quite literally, undergraduate projects.
I don't see people upvoting "My First Stack" or "Check out this Pet Shop I wrote in J2EE". Computational intelligence is not some land of super wise geniuses that mere mortals cannot ascend to. It's all pretty straightforward.
Eg, my Ant Colony Optimiser for a Travelling Salesman Problem:
https://github.com/jchester/ruby-ry48p-aco https://github.com/jchester/ruby-ry48p-aco
and my Genetic Algorithm for optimising Ackley's Function:
https://github.com/jchester/ruby-ackley-genetic-algorithm https://github.com/jchester/ruby-ackley-genetic-algorithm
- neurotech1 14y agoGive the guy a little more credit. Most Machine Learning/Neural Network code is in python or C/C++ not PHP. Something like this would be a good starting point if moving to Python is not an option.
- jacques_chester 14y agoA perceptron is 1. simple to code up and 2. not very useful.[1] You still need to pick the sort of neural network you're building (there are literally hundreds in the literature) before anything of consequence is going to happen. The reason to use C/C++ libraries is that neural nets are slow. Slow slow slow. Computational intelligence / nature inspired computing is extremely cool, but the smart thing is to avoid it like the plague unless it is the only workable alternative. [1] To the point that neural network research came to a screeching halt for some time when it was proved that a perceptron cannot perform a XOR on inputs.
- illicium 14y ago>a perceptron cannot perform a XOR on inputs Multi-layer perceptron networks, however, can.
- jacques_chester 14y agoRight. According to my old notes you need: * at least one layer to handle linearly separable functions, * at least one hidden layer to handle continuous functions, and * at least two hidden layers to solve continuous and discontinuous functions.
- jacquesm 14y ago> This makes me think I should upload my random computational intelligence assignments and get sweet, delicious karma. Does karma matter that much to you? > Edit: OK, I was a bit snarky, but I'm serious. Well, what's stopping you? The fear of getting people to hate at you like this in comments? Really, you can do much better than this. > I see these things on HN all the time. These are, quite literally, undergraduate projects. They are people taking steps on the ladder to proficiency, as such they should be encouraged rather than be beaten down. > I don't see people upvoting "My First Stack" or "Check out this Pet Shop I wrote in J2EE". I don't see people submitting those either, which is one reason they're not being voted up. Once upon a time you too were a beginner. And in the eyes of many you probably still are. > Computational intelligence is not some land of super wise geniuses that mere mortals cannot ascend to. It's all pretty straightforward. Exactly! And what better way to demonstrate that by showing it, in a language that is generally despised. > Eg, my Ant Colony Optimiser for a Travelling Salesman Problem: > https://github.com/jchester/ruby-ry48p-aco https://github.com/jchester/ruby-ry48p-aco > and my Genetic Algorithm for optimising Ackley's Function: > https://github.com/jchester/ruby-ackley-genetic-algorithm https://github.com/jchester/ruby-ackley-genetic-algorithm That's pretty neat code ;)
- jacques_chester 14y ago> Does karma matter that much to you? Revealed preferences suggest that it does :( > That's pretty neat code ;) Flattery will get you karma beyond your wildest dreams (for wildest dreams = 1).
- jacquesm 14y ago> Flattery will get you karma beyond your wildest dreams (for wildest dreams = 1). It's encouragement, not flattery.