@Wellenreiter I added more samples last night. some are more dubious than others (as is the nature of data mining) I added some graphics to … favorite.ico and a big one int how to page (got them from shutter stock and edited them up in gimp). I haven’t advertised anywhere yet, i will definitely look in to doing it those places. i did join the 2 biggest facebook groups just so i share a link there but i think i dont think i’m quite what people expect in those groups.
@AcidD I actually haven’t put my code in github at all. I do stock analysis as well for personal investment and of course work on kaggle contests from time to time. It all uses some version of this code. so pretty much everything i have I kinda want to keep to myself till i get some sort of payout from it. Then well anyone can have the code. Is that greedy? its been a hobby since the netflix contest years ago. I’d just like to see something monetary wise before i give it to the world.
I also have a TSNE implementation that is linear runtime, i might add to the website some day. As well as my genetic stuff that does all kinds of magical things. But i’m about to retool that to work on making new features instead of solving the big problem (i think it’ll do a better job of that)
There will probably be 1 more significant update to the site where i add 2 or 3 more samples and add the ability to have features that are categorical via an optional 2nd header row (which can either be split automatically for you via one-hot-encoding, or compiled down to averages solution value for that particular categorical value, your choice)
Just when I get an Intel PC for the first time, it turns out their speed increase over AMD is due to and insecure processors’ out-of-order execution to read arbitrary kernel-memory location.
Spectre and Meltdown are side-channel attacks which deduce the contents of a memory location which should not normally be accessible by using timing to observe whether another, accessible, location is present in the cache.
These caches can include personal data and passwords.
Test your Linux system is patched for Meltdown (or not) Spectre . : https://github.com/speed47/spectre-meltdown-checker
It’s already been the big thing for a couple years.
I can’t tell you specifics of what we’re doing with it, but we’ve had an 8x volta 80Gbit nvlink system on order for a couple months now.
Imagine what 40,000 cuda cores, 5,000 tensor cores, and HBM crazy-fast memory all ganged up in one box can do these days. Distributed systems have been designed to allow several multi-gpu boxes to be connected. That’s the reason ML/DL has taken off so quickly; this kind of raw power was practically unthinkable a mere half dozen years ago, and is making models only dreamed of before workable today. But then we all knew where gpgpu was headed anyway…
I had picked up a book on neural networks twenty years ago, and wondered then just what good that would ever serve in our lifetime. Who knew.
(btw, quantum computing is at the same stage now that nn’s were back then… :))
I figured out everything i needed. the short version is I’m going to use/got working a sample of openCL . (I tested a local build on different machines each with 2 different graphic cards and all seems to work well)
The long version check out this link https://www.codeproject.com/Articles/1116907/How-to-Use-Your-GPU-in-NET you will also need to install intel’s opencl and the latest graphics drivers on the machines you are working on. The final code seems to actually be a script that is compiled at runtime and sent in to the graphics cards. it worked on nvidia and 7990 (2 gpus) and R9 series amds so… seems to work quite well.
