|
| 21 Oct 2013 04:19 PM |
My concept is that if we can create a simple neural network, and then implement MD5 or SHA1 hash checking into an array, Lua can be trained to simply propagate said network and create more RAM available for physics use. I guess what I'm trying to say is that by adjusting the weights, our array can become a global variable and then use some form of instancer to emulate genes. Just a thought. You would be of help if you have a suggestion for making the script a bit more efficient when it comes to the XORing, however.
flip={} function propagate() xor=true
I'm new to Lua, but I think I have a pretty firm grasp on the context. Thanks for reading! |
|
|
| Report Abuse |
|
|
|
| 21 Oct 2013 05:10 PM |
| Oh, by the way, it's also acceptable to use a lesser checksum. |
|
|
| Report Abuse |
|
|
|
| 21 Oct 2013 06:58 PM |
| Are you talking Lua, or RBLX.Lua? These days, there are so many different additions of Lua, such as Love, and some other stuff. |
|
|
| Report Abuse |
|
|
cntkillme
|
  |
| Joined: 07 Apr 2008 |
| Total Posts: 44956 |
|
|
| 21 Oct 2013 07:24 PM |
^ that's not how it works, out please k?
Love2D is just a game engine I believe |
|
|
| Report Abuse |
|
|
|
| 21 Oct 2013 07:30 PM |
| No need to throw a "u" in that name of yours, Cnt. Not everyone is as talented as you at this. |
|
|
| Report Abuse |
|
|
cntkillme
|
  |
| Joined: 07 Apr 2008 |
| Total Posts: 44956 |
|
|
| 21 Oct 2013 07:32 PM |
| I'm not, he should just say "I think" or something so we know that he isn't pointing it out as facts. |
|
|
| Report Abuse |
|
|
Oysi
|
  |
| Joined: 06 Jul 2009 |
| Total Posts: 9058 |
|
| |
|
cntkillme
|
  |
| Joined: 07 Apr 2008 |
| Total Posts: 44956 |
|
|
| 21 Oct 2013 08:39 PM |
No, love is not built in Lua. I hope you aren't joking |
|
|
| Report Abuse |
|
|
Bobobob12
|
  |
| Joined: 23 Jan 2008 |
| Total Posts: 5350 |
|
|
| 21 Oct 2013 08:59 PM |
| Let's not forget to analyze ALL of your vectors shall we? It's common fact that SHA1 is a 64 bit method of key seeded encryption. The neural network proposition is just madness; it would be much simpler to begin with the simulation of the exponentially growing lower function tissue and expand from there. The theoretical instancer would actually and eventually generate itself in time. This is not to ignore the other major flaws in your algorithm. Without proper constructors and deconstructors within the definition of your MD5 propagation, the manually allocated RAM will leak over and corrupt your data, not to mention overflow the buffered cache, truncating the array terminator. Also, if a bit-to-BCD (and vice versa) isn't implemented, you certainly can't hope to generate a true gnome, as the decimal system is surprisingly important in the process of higher organs and brain function. A novice mistake, I certainly hope you'll be more careful in the future. |
|
|
| Report Abuse |
|
|
|
| 22 Oct 2013 04:22 PM |
@Bobobob12
Again, I am just learning Lua. I can indeed now see the error that you have showed me. Luckily, I saved a backup of the neural network's genome on a Solid State Drive (SSD) that I scripted. I am currently working on a patch for the MD5 propagation network. I did create a highly advanced RNG that will seed the weights when it comes to the neural network. Also, I hate to point out a mistake of yours, however, the way that I have implemented my genome processor function (if you care to check out the original script again) will show that I used an alternative method and thus the decimal system is not needed. I used a binary-only approach, simply simulating around 50,000 neurons. It does also grow new neurons over time, but the instancer isn't yet calibrated for this. Once I get it, I'll post an updated code. For now, here's a new piece I added which can handle a lot of the problems you were speaking of.
fixes={0,0,1,1,0} function apply_neural() apply_neural() end apply_neural() for i=1,100 fix(i) end
As you can see, it does contain a genome in it that can mutate at a rate of 1/32 of the regular neural network, allowing it to patch any bugs we accidentally create while writing this (over time, of course.) |
|
|
| Report Abuse |
|
|
|
| 22 Oct 2013 04:24 PM |
This Thread ,Is Now ,Mine .
~AbsoluteLOL. |
|
|
| Report Abuse |
|
|
|
| 22 Oct 2013 04:32 PM |
GTFO, moron.
(yay floodcheck) |
|
|
| Report Abuse |
|
|
|
| 23 Oct 2013 06:59 AM |
| I don't get what you're saying, the only time I heard "propagate and XOR" together is in adders, and aside from that, I also believe that something is only as strong as the underlying machine. |
|
|
| Report Abuse |
|
|
| |
|
|
| 23 Oct 2013 08:52 AM |
| Radio, what to him or to me? |
|
|
| Report Abuse |
|
|