BETA - Neural Network in Java. Numbers Predictor

Reply #38

osmannica2001 — Thu, 28 Sep 2017 03:34:17 GMT

Todd,

Neural networks work at the decimal level of math. Why? because in order to find the right weight of values that may represent an accurate output, we must first find N tangent lines which present the instantaneous rate of change of the function at a certain one point. The slope of the tangent line at a point on the function is equal to the derivative of the function at the same point, and when trying to find the limit at which a rate of change is performing it is necessary to zoom in at... [ More ]

Reply #37

osmannica2001 — Thu, 28 Sep 2017 03:21:16 GMT

Hi,

Thanks for your constructive feedback.

Recall the following laws of logic!

Given any statement variables p,q, and r, a tautology t and a contradiction c, the following logical equivalences

hold.

p q q p p q q p

(p q) r p (q r) (p q) r p (q r)

p (q r) (p q) (p r) p (q r) (p q) (p r)

p t p p c p

p p t p p c

( p) p

p p p p p p

p t t p c c

(p q) p q (p q) p q

p (p q) p p (p q) p

t c c t

Contradictions and equiva... [ More ]

Reply #36

osmannica2001 — Thu, 28 Sep 2017 02:59:13 GMT

lol not at all - at the end of the day its just a Game. Not many people believe in trends, but those who do and persist in finding answers will succeed.

Wealth is rarely shared - rather it is an individual process. So, time to keep it up

Reply #35

lakerben — Fri, 15 Sep 2017 14:07:17 GMT

I'm always hitting 3 of 5 in the NM Roadrunnercash 5/39.

I don't use software I use number groups ,factoring etc. I even use Dr M number groupings 2509,358,6401 etc. It does work when I anticipate what's going to hit. Surprise,surprise like Gomer Pyle use to say!

If i spend more time I'm sure I will hit the 5 numbers. I hit again last night.

Reply #34

RL-RANDOMLOGIC — Tue, 12 Sep 2017 06:29:07 GMT

Ricklou

I was not trying to be negative, patterns do exist and not just a few. What I say is that these patterns

are not always trustworthy. The patterns we so often see, are too often just a product of random. A

real pattern will repeat and that's how we know it's real.

A good NN will build the best solution based on the games history but much of the data it uses for analysis

is random which gives false positives.

It's a fine line and before any real progress can be made in... [ More ]

Reply #33

MillionsWanted — Fri, 08 Sep 2017 14:32:02 GMT

I hope Osmannica hasn't given up on his ANN after these negative posts. Looking forward to next update.

Reply #32

Ricklou — Mon, 04 Sep 2017 13:04:49 GMT

Hi there lucky1day

Look as I sceptic myself for years I got the answer from a reliable source in the industry how lotteries really work and to be honest with you if you think about it logically all the puzzle pieces will fall into place. You should understand a business main goal is to be profitable -if they are not profitable they cannot go on. Lotteries around the world will only let jackpots go if they have profit and they can let the last jackpot go to a winner. That make 100% business se... [ More ]

Reply #31

Ricklou — Mon, 04 Sep 2017 12:40:52 GMT

Hi there RL

If you analyze lottery numbers on a regular basis you will pick up stuff that is infact patterns -you should know with your good lottery software you are writing -much respect to you as I am doing the same. Like I have mentioned you will start seeing stuff with years and years of analyzing numbers. Will give you another example I refer to consecutive skips next to each other -this patern DO NOT exist on your MM or PB often(Your Jackpot LOTTERIES is the ELITE OF THE ELITE LOT... [ More ]

Reply #30

thamizhpayan — Thu, 31 Aug 2017 18:31:31 GMT

I agree browser showing message about outdated version of Java.And some time tell that that code could use a vulnerability in the outdated JRE that you are nunning and compromise something locally...

Give some references to most website no longer allow Java either for same reason

Every software has vulnerability that will be exploited - starting from OS...

Any JavaScript popup with a browser vulnerability breach can infect your computer so will you endup telling not to u... [ More ]

Reply #29

thamizhpayan — Thu, 31 Aug 2017 18:22:34 GMT

C and C++ are natively compiled and native binaries are executed directly.

Java code on the other hand is compiled into a byte code which runs on a Java Virtual Machine(JVM)

So where is this security issue peeping here?

Why would Banks and Insurance companies use java in their Web/GUI fronted and backend

Reply #28

thamizhpayan — Thu, 31 Aug 2017 18:18:17 GMT

No I strongly disagree.

Java Applets having security issues is different and I don't see any applets being used nowadays.

Where as full blown applications written in Java(either Web ui frameworks or swing or console apps) are no less secure than code written in any other language.

I would like to mention that lot of information stored and used by insurance companies, banks and other companies are hosted and run java enterprise applications.

As a developer, the onl... [ More ]

Reply #27

MonEl — Thu, 31 Aug 2017 17:22:52 GMT

I don't know, but somebody said somewhere that C is similar to Java, I don't like to have Java in my computer either.

Reply #26

Kosh — Thu, 31 Aug 2017 03:11:00 GMT

Dear Osmanica2001,

I wish you would not write the software in Java. Most people have disabled Java on their computer because that programming language makes computers vulnerable to viruses and malware. Most websites no longer allow Java either for the same reason.

You should write your software in the same programming language as Stoopendaal's Neural Net Lottery Picker (NNLP) or Perceplotron instead or as an additonal version for those who do not want to have Java on their computer.... [ More ]

Reply #25

lucky1day — Wed, 30 Aug 2017 21:07:06 GMT

Well, it amazes me too! People don't want to believe that their being screwed. When I play, I play for the second tier prizes simply because they have to award those to keep interest in the game.

As for the actual topic. I believe machine language is not quite there yet to predict dissimilar inputs and outputs. I've actually written my own code in python as well as mathlab and the best I've been able to do on a consistent basis is 3 out of 50 numbers. That leaves quite a mountain to climb.... [ More ]

Reply #24

lakerben — Wed, 30 Aug 2017 15:13:58 GMT

Rl which program is this and can I get the download

Reply #23

lakerben — Wed, 30 Aug 2017 15:10:01 GMT

Getting back to the system how is that going? Too much verbage because any major games have repeat digits.

Reply #22

RL-RANDOMLOGIC — Wed, 30 Aug 2017 12:46:01 GMT

Rickiou

True random does not exist, unfortunately, sufficiently random does. Lotteries don't need to cheat

the game odds are designed to earn profits. The house always wins, even when there is a big payout.

Don't forget, they pocket around 50% of sales just for running the game. The only cheating I have seen

has been by insiders hoping to score big. There are ways to play smarter but these fall short of handing

us a JP. Tonight I hope to win the MO 6-44 game. I plan to say a littl... [ More ]

Reply #21

Ricklou — Tue, 29 Aug 2017 20:14:33 GMT

Just want to add something to the mix apologies I have neglect to add it above explaining patterns.....if the winning jackpot numbers get properly analyzed via skip patterns you will discover randomness is a myth....Example how appropriately for skip 11 to just fall with skip 12(Perfectly chosen and grouped together)......what are the odds of that ever happening in true random events. There is skill and art in playing the lottery and there is a blueprint.....very little people have cracked that

Reply #20

Ricklou — Tue, 29 Aug 2017 19:23:27 GMT

Patterns exist like odds evens etc........but I agree with you 3000% ANY LOTTERY WORLDWIDE are FULLY CONTROLLED....it amazes me people who still believes ping pong balls fly and just get randomly chosen.......Sorry its not the case a jackpot ONLY goes when the lottery is in profit.....who will give away a golden egg when you making a loss ???? Exactly!! People need to break that type of stereotype of thinking as they don't know how the industry work. This only applies to Million Dollar Jackpot P... [ More ]

Reply #19

Todd — Tue, 29 Aug 2017 14:36:26 GMT

osmannica2001, I was surprised to see you write that floating point operations are quicker than integer operations. In my programming experience I have never seen that. A computer does not see a big number like 1000 (or 1,000,000,000) as more difficult or time-consuming to process than a fractional number like 0.054. In fact, the opposite is generally true, and it's due to the fact that CPUs are built for integer operations at their core.

Reply #18

thamizhpayan — Tue, 29 Aug 2017 14:22:54 GMT

RL, Spot on. Even in LP, discussions and tools have been posted even 10 years back.

NN, AI, Machine learning - they learn patterns and predict(essentially categorize -correct me if I am wrong). Even Stoopendal in his NN tools mentioned that he doesn't think that NN can predict winner but can surprise at times.

It's fun to try, choose, play(for real if you have money or on paper if you don't want to spend) and then loose.

As we all know there are too many variables(especially b... [ More ]

Reply #17

RL-RANDOMLOGIC — Tue, 29 Aug 2017 13:53:58 GMT

Spot9

You can't shorten the odds for the game, you can however reduce the field of choices one has to make

or work with data that might be more predictable.

In the post above you wrote even + even=0, odd + odd= 0, even + odd= 0 and odd+ even=0.

It should be even + even=1, odd + odd= 1 and even + odd= 0 and odd+ even=0. or something similar.

All four choices can't equal the same value. Here is a pic of a old P-3 program that does close to the same

thing. This setup produced... [ More ]

Reply #16

Spot9 — Tue, 29 Aug 2017 08:16:00 GMT

Hi Igamble,

The first choice for odd/even digits for the three positions has only 8 possible choices.

For the second stage you would then choose odd/even values for the pairs.

This is a bit like Rolf Speidel's two stage fish net system, but I don't know

if you would gain much from it.

http://www.wettsysteme.de/

Spot9

Reply #15

Igamble — Tue, 29 Aug 2017 02:41:32 GMT

Is not as simple as your friend had described you.

Just starting from a 101 pattern you have xxx Straight combos. THEN starts the real work...the next step when you get another pattern, again is part of many patterns of anywhere of X-X combos. And that is nothing gained until you spend some serious work stuyding the skips and history data, which depending of what state you play may or not be rigged....

So, it will come down to your playing budget, and a chance that those straight sets yo... [ More ]

Reply #14

RL-RANDOMLOGIC — Sun, 27 Aug 2017 21:27:08 GMT

Spot9

It's all about the data, if we can find ways of converting the data into other formats where some of the randomness

goes away then we can make better choices. We may not be able to zero in on the exact digit, number etc... but

anytime we can narrow the field of choices it's a step in the right direction.

RL

Reply #13

RL-RANDOMLOGIC — Sun, 27 Aug 2017 21:18:56 GMT

I have written many such programs and none of them work when applied to random data. Sure, it's possible

to get a solution just don't expect the numbers drawn to have any connection to that solution. This stuff is

not new here, it's like watching reruns for the 10th time. Bayes, NN, AI, machine learning, adaptive boost

etc... have all been tried and failed. These tools work fine except when applied to random prediction and

the main reason they fail is, well, the lottery is random. T... [ More ]

Reply #12

thamizhpayan — Sun, 27 Aug 2017 05:36:09 GMT

Thanks, I will try it and let you know.

Good Luck.

Reply #11

Spot9 — Sun, 27 Aug 2017 01:10:26 GMT

P.S. I think the code for even/odd and odd/even pairs should be 1?

Spot9

Reply #10

Spot9 — Sun, 27 Aug 2017 00:45:13 GMT

Hi RL,

I agree in reducing the lottery to limited values to try to use/predict.

This is a post I saw on LP about Binary Lotto:

A friend of mine wins consistently by using what he calls binary codes for the P3. What he seems to do is reduce each position of the pick 3 numbers to either 0 or 1, depending on whether they are even or odd. The numbers 0,2,4,6,8, are even=0, and 1,3,5,7,9, are odd=1. He does this to each of the positions. He them seems to put the first 2 positions together a... [ More ]

Reply #9

tvpventures — Sat, 26 Aug 2017 20:40:08 GMT

Hi, ocmannica2001, thank you for sharing your tool. I want to try this over the weekend or Monday.

I like what you said in your post: In my subjective opinion, Randomness doesn't exist ....... Me too, I think about this a lot.

My favorite saying from Karl Jung:

In all chaos there is a cosmos, in all disorder a secret order .

I am very sure it also in a lottery randomness, there is possibly hidden patterns that our brains cant grasp it yet (some might did, we never know..:-) Anyway,... [ More ]

Reply #8

osmannica2001 — Sat, 26 Aug 2017 20:18:32 GMT

The idea of machine learning is to help you decrease the marginal error with respect to an output of events given the likelihood of another event occurring. The law of average indicates that the principle that supposes most future events are likely to balance any past deviation from a presumed average - something that machine learning allows to visualize easier, rather than manually filling pages and pages of manual statistical calculations.

In my subjective opinion, Randomness doesn't exist.... [ More ]

Reply #7

osmannica2001 — Sat, 26 Aug 2017 20:07:27 GMT

Hi, I am not sure what operating system you may be using, but I guess just make sure that your Java software is working correctly. I'd say reinstalling Java may help.

Reply #6

osmannica2001 — Sat, 26 Aug 2017 20:05:30 GMT

It is not built on top of any frameworks. The code is similar by the one built by this guy Kunuk Nykjaer - http://bit.ly/2wel0ch.

Reply #5

RL-RANDOMLOGIC — Sat, 26 Aug 2017 18:50:10 GMT

HI Guys

I don't want to sound like Johnny rain-cloud but these type tools won't predict random. The reason IMHO

is that the field of choices is too wide and the next drawing is random. If your serious about building a good

prediction tool then you need to start with data that is somewhat more predictable. Example, will the draw

contain more even or more odd numbers. There is no way to know for sure but given the output is limited to

two choices a tool such as this might be better... [ More ]

Reply #4

jlg69 — Fri, 25 Aug 2017 18:47:34 GMT

Good evening,

Thanks for your work, unfortunately i get an issue when I run your jar-file

impossible de trouver ou charger la classe principale machinelearningtest.MachineLearningTest

(impossible to find or load the main class for ....)

Do you have a clue ? As a newbie in java class system

Thanks

Reply #3

thamizhpayan — Fri, 25 Aug 2017 11:58:55 GMT

Is the neural network core of your program written from scratch or does it use Neuroph or Weka or any other known framework?

I experimented with some of them.

I would say if a tool allows back testing huge parameter sets, then that would help a lot (like batch mode). Stoopendal's tools are the only similar tool like yours.

Thanks and Good Luck

Reply #2

osmannica2001 — Thu, 24 Aug 2017 06:28:32 GMT

It goes in bottom.

Reply #1

MillionsWanted — Thu, 24 Aug 2017 05:51:32 GMT

Thanks.

I'll check it out.

Latest draw on top or bottom

BETA - Neural Network in Java. Numbers Predictor

osmannica2001 — Thu, 24 Aug 2017 04:30:43 GMT

Hi all,

In the past I have shared code related to machine learning prediction, but it seems as it may be quite complex to use for practice. This time, I got my hands dirty with Java and was able to build a graphical interface software that uses Neural Network to predict data.

Do know that this is a beta testing for anyone interested in playing with different settings of the software, this is to try to predict future lottery numbers based on past history data.

In order to run the s... [ More ]