Quick Links   You last visited June 27, 2017, 1:10 am All times shown are
Eastern Time (GMT-5:00) | Neural Net Lottery PickerNorway Member #9517 December 10, 2004 1329 Posts Offline | | Posted: October 31, 2015, 8:08 am - IP Logged | |
I've been using perceplotron to pick numbers for pick 4 on this board's prediction section for the past week. I've gotten lots of 3 number matches, but no 4 digit matches. Sooooo Close! If you managed to make a win once every third month you would still make good profits? | | |
Thread Starter Netherlands Member #3476 January 24, 2004 212 Posts Offline | | Posted: October 31, 2015, 3:47 pm - IP Logged | |
Hi lottolot, If you want to use the 'restart training' function you have to activate (checkbox) this function before you start 'training' the ANN. When you check this function while the training is already running it will be skipped. In the next version -I'm working on it  - of NNLP I'll disable the checkbox if the training is already running to avoid confusion . Cheers! | | |
Thread Starter Netherlands Member #3476 January 24, 2004 212 Posts Offline | | Posted: October 31, 2015, 3:56 pm - IP Logged | |
Hi ArizonaDream, NNLP was not designed for digit games like pick 3 etc. Maybe I'll implement this option in a future version of NNLP. Cheers! | | |
Tanhauser Gates Holy See (Vatican City State) Member #139281 February 18, 2013 200 Posts Offline | | Posted: October 31, 2015, 6:19 pm - IP Logged | |
Hi lottolot, If you want to use the 'restart training' function you have to activate (checkbox) this function before you start 'training' the ANN. When you check this function while the training is already running it will be skipped. In the next version -I'm working on it - of NNLP I'll disable the checkbox if the training is already running to avoid confusion . Cheers! Thanks for your answer,Stoopendaal. Regards 
| | |
Tanhauser Gates Holy See (Vatican City State) Member #139281 February 18, 2013 200 Posts Offline | | Posted: November 1, 2015, 4:13 pm - IP Logged | |
Hi lottolot, If you want to use the 'restart training' function you have to activate (checkbox) this function before you start 'training' the ANN. When you check this function while the training is already running it will be skipped. In the next version -I'm working on it - of NNLP I'll disable the checkbox if the training is already running to avoid confusion . Cheers! Stoopendaal,thanks for your suggestion. It's Works very well.!!!. Regards . | | |
United States Member #169277 October 10, 2015 631 Posts Offline | | Posted: November 1, 2015, 4:41 pm - IP Logged | |
It would be awesome if you did make a pick 3/4 version.. | | |
Australia Member #166177 May 8, 2015 65 Posts Offline | | Posted: November 3, 2015, 5:06 am - IP Logged | |
Hi ArizonaDream, Try a program called T-LEX from Randomlogic - it seems to work very well with pick 3/4 digits. Cheers, Gambler667 | | |
United States Member #147790 October 13, 2013 2111 Posts Offline | | Posted: November 4, 2015, 2:50 pm - IP Logged | |
Hi ArizonaDream, NNLP was not designed for digit games like pick 3 etc. Maybe I'll implement this option in a future version of NNLP. Cheers! Stoopendaal! Thank you FYI P3 does work (one draw state). I tested the date range below >>> RED Start of Test (did not enter draws after 10/13/2015) -- Results in Green Tue-Nov-03-2015 E 721
Mon-Nov-02-2015 E 753
Sun-Nov-01-2015 E 142
Sat-Oct-31-2015 E 917
Fri-Oct-30-2015 E 221
Thu-Oct-29-2015 E 453 Str8
Wed-Oct-28-2015 E 617
Tue-Oct-27-2015 E 120 Box
Mon-Oct-26-2015 E 988
Sun-Oct-25-2015 E 496
Sat-Oct-24-2015 E 078
Fri-Oct-23-2015 E 776
Thu-Oct-22-2015 E 418
Wed-Oct-21-2015 E 034
Tue-Oct-20-2015 E 082
Mon-Oct-19-2015 E 715
Sun-Oct-18-2015 E 892
Sat-Oct-17-2015 E 272
Fri-Oct-16-2015 E 668
Thu-Oct-15-2015 E 834
Wed-Oct-14-2015 E 787
Tue-Oct-13-2015 E 267
Mon-Oct-12-2015 E 766
Sun-Oct-11-2015 E 535
Sat-Oct-10-2015 E 871
Fri-Oct-09-2015 E 764
Data used : Note Draw 267 is where I started (Red) NNLP --- As I was creating this note NNLP gave also 672 (Blue Below) I suspect a repeat on the way 04-05-03-;18.498281989971;0.608248008463176;0.482904599267874;11:25:04.3469822;Learning Rate:0.601283| Momentum: 1E-05| Noise:0.05|Threshold Error:0|Hidden Neurons:50|Test Data Range:50;)
02-01-00-;18.4410316450437;0.607306045500021;0.494342936772287;11:25:19.9178728;Learning Rate:0.601283| Momentum: 1E-05| Noise:0.05|Threshold Error:0|Hidden Neurons:50|Test Data Range:50;)
06-07-02-;18.2765794945576;0.604592085534663;0.479546016689102;11:26:49.2469821;Learning Rate:0.601283| Momentum: 1E-05| Noise:0.05|Threshold Error:0|Hidden Neurons:50|Test Data Range:50;)
02-01-00-;18.2639188974999;0.604382641999254;0.480074926999595;11:27:46.3472480;Learning Rate:0.601283| Momentum: 1E-05| Noise:0.05|Threshold Error:0|Hidden Neurons:50|Test Data Range:50;) 
NOTE: I started testing 4 days earlier and found 453 and 721 (latest draw) came out a lot using several different parameter levels. I have my reasons but I found 14 - 18 days early is optimal for 1-draw. Two draw states could bump to around 8-14 days (worth the note) 1.61803398875 | | |
United States Member #147790 October 13, 2013 2111 Posts Offline | | Posted: November 4, 2015, 3:00 pm - IP Logged | |
Maybe it's just my PC so this is just in case you are not aware >>> version Percept-Lotron 1.3 the "Start Bactesting ANN" button does not respond. Also I get this error when trying to make the charts <<< I can live the way it is no worries, no pressure 
Thanks again for your hard work 1.61803398875 | | |
Tanhauser Gates Holy See (Vatican City State) Member #139281 February 18, 2013 200 Posts Offline | | Posted: November 4, 2015, 3:49 pm - IP Logged | |
Today,another 3/6. I´m VERY happy with this program. Before NNLP,I spent more than 1 hour,working for get my numbers. Now,I spend a a few SECONDS. And the results are very interesting and promising... Thanks again Stoopendaal. | | |
Greensboro, North Carolina United States Member #1616 June 5, 2003 1301 Posts Offline | | Posted: November 8, 2015, 9:34 am - IP Logged | |
Today,another 3/6. I´m VERY happy with this program. Before NNLP,I spent more than 1 hour,working for get my numbers. Now,I spend a a few SECONDS. And the results are very interesting and promising... Thanks again Stoopendaal. If you don't mind me asking, what are you using for settings ? Mine are either too much or not enough to get better than a match 2. | | |
Norway Member #9517 December 10, 2004 1329 Posts Offline | | Posted: November 8, 2015, 2:08 pm - IP Logged | |
Perhaps Stoopendal can explain the Ranking values used on Percep-Lotron? Typical Ranking values are 388,XX , 105,XX , 99,XX , 98,XX on the right side of the Predictions numbers(Top-Ranking Numbers) list. Does the Ranking value indicate how good the numbers are? If all the Ranking values are low, like under 110, is the prediction badder? | | |
Tanhauser Gates Holy See (Vatican City State) Member #139281 February 18, 2013 200 Posts Offline | | Posted: November 8, 2015, 5:45 pm - IP Logged | |
If you don't mind me asking, what are you using for settings ? Mine are either too much or not enough to get better than a match 2. Hello reddog: I think every game it has it´s own settings,but anyway,this is my personal 6/49 settings: Learning:0,174848 Noise:0,176 Threshold:0,053731 Neurons:77 Training data:27 The rest of settings:Nothing. I hope this help!!! Best regards | | |
Greensboro, North Carolina United States Member #1616 June 5, 2003 1301 Posts Offline | | Posted: November 8, 2015, 6:36 pm - IP Logged | |
Hello reddog: I think every game it has it´s own settings,but anyway,this is my personal 6/49 settings: Learning:0,174848 Noise:0,176 Threshold:0,053731 Neurons:77 Training data:27 The rest of settings:Nothing. I hope this help!!! Best regards Thank you, I will try these and see what it does for 5/39.  | | |
Tanhauser Gates Holy See (Vatican City State) Member #139281 February 18, 2013 200 Posts Offline | | Posted: November 13, 2015, 3:40 pm - IP Logged | |
After some days with no wins...today another 3/6. (Eight bets played,and one winning bet). I hope soon a 5/6 prize... Regards | | |
|