Search

Neural Net Lottery PickerReply
Millionswanted, jlg69, The ANN implementation in version 1.5 of perceplotron is (much) slower than the ANN I used in the previous versions of perceplotron. I added an extra stopping condition in the latest version (1.6) to speed up the training hopefully without loosing performance. This extra condition, mean cross entropy error, checks the mcee of the training data (lottery draws) when training the network. When the found mcee value is smaller than the mcee value you entered the t
Oct 24, 2015, 8:58 am - stoopendaal - Mathematics Forum

RL's T-Lex program download, free softwareReply
Hi RL, try to open program but gave me an error saying SDL_image.dll was not found..:( not sure whats going on here. Windows xp pro.. Thanks for your time :)
Sep 25, 2015, 12:38 am - Eddie233 - Lottery Systems Forum

Lottery Bible ~ Large PrintReply
Thanks Harvest Moon, Won't use the Lady Einstein again unless you give the ok. I saw one error on the working numbers (501) and just went to my book for reference wasn't aware of pg 2. I had posted to Denise from Florida and told her about pg 1 and wanted to tell her some of the #'s I had just posted look good 886 was related to her previous winning #'s 049/448 and I also 302 which fell next in Florida, but She never posted back. Also posted to Fitzmarr from New York 979 combo fell and I jump on
Sep 12, 2015, 5:54 pm - DIAMONDRAY - Lottery Systems Forum

Slider downloadReply
Adam, if I choose 15 samples and up the error doesn't show. ONLY when I use 10 samples.
Sep 9, 2015, 11:48 am - adulane62 - Lottery Systems Forum

Slider downloadReply
The problem with that error is apparently when your sample size is exactly the same value as your recalc let alone higher. So for example when you use a sample size of 50 then you should put in Recalc box the value of 49 and not 50. Then it works ok. Adam
Sep 9, 2015, 11:28 am - adamcustom - Lottery Systems Forum

Slider downloadReply
Rl, It does not matter which game is loaded. If I set it to 10 past draws it still shows the error in my previous post. Pushing the no button shuts down the application, and the yes button will let you continue to the next step. However, is the results it shows for the past 10 draws or something else? Also does setting the regression to 5 max help or hinder the selection process? Thanks.
Sep 9, 2015, 10:49 am - adulane62 - Lottery Systems Forum

Georgia Team Players Post Pic 4"S HereReply
Your 4961 reminds me of a number I meant to play but forgot: 4960 Hopefully she will over look the error and give me 9960 instead !
Sep 7, 2015, 11:25 pm - KAL035 - Pick 4 Forum

Neural Net Lottery PickerReply
Just noticed: I'm no longer able to set Threshold/Error at 0.01. Instead I must set it to 0.010048. Not sure if there are any practical differences.
Sep 5, 2015, 8:51 am - MillionsWanted - Mathematics Forum

Neural Net Lottery PickerReply
hi winsumloosesum, In the screen shot below does it make a difference if there are dashes after the last entry. (see arrows) == no, it's not allowed to add a dash, space, etc. after the last entry. A error message will be shown and the draw results will not be loaded for training.
Aug 23, 2015, 2:21 pm - stoopendaal - Mathematics Forum

Looking for Excel FormulaReply
No, the formula still gives an error message. =IF(COLUMNS($R10:R10) =COUNTIFS($B$2:$B$5316, =1 ),COUNTIFS($B$2:$B$5316, =50 ),COUNTIFS($B$2:$B$5316, =1 ),COUNTIFS($B$2:$B$5316, =50 ),ROW($B$2:$B$5316)-ROW($B$2)),COLUMNS($R10:R10))
Aug 20, 2015, 3:07 pm - CARBOB - Lottery Systems Forum