Neural networksPrevious TopicNext Topic

Has anyone experimented with neural networks?

If so, what was your experience.  I am just now entering the world of neural networks and have idea whether my time will be wasted or well spent.  I am hoping a LP member will offer a few suggestions.

Thamks

If the lottery is not random enough they might be able to find a pattern if not, it's just like guessing.

The network type that seems to work best is an algorithm from GMDH which is used by aiLotto and Ward Systems Group.
aiLotto claim their lottery software it is able to predict one line with 3/6 in every fourth draw. I don't own the software myself.

Interesting. I was just reading about bee and ant algorithms the other day and wondering how they could be applied.

Isn't it funny how we try to take everything and see how it fits into the lottery?

yes I do use them.

Its not worth time for daily games.

Look at it like this "reverse node" back tracking.

after all you set the rules as were to look for in history.

Precisely my point only I am not being funny.  As there is no definitive answer to randomness at this point in time, we have to look at everything and see if the principles involved can be applied to the lottory.

Case in point:  For hundreds of years no  mathematician could chart for you the path of comets. Then along came Mr. Newton and his Calculus.  Edmund Halley, using Calculus, then determined that the comet named after him returned to this vicinity of the universe every 75 years and has been doing so for centuries.  He took something and applied it elsewhere.   Should we do no less?

While neural networks may not be the holy grail we seek, we should not turn our intelligence away from its potential benefits.  Until another of Mr. Newton's calibre comes along and exposes the mystery of randomness, we of lesser intellectual stength should consider every possiblity.  At the very least we should refrain from denigrating those who are trying ( tho I do not consider your reply to be denigrating ).

Maybe this will be of interest:

http://www.lotterypost.com/thread/113989/391541

AAWW Gasmeterguy I meant no harm. I don't even know what neural networks are. My reply was in reference to myself....... as the more theories and ideas that are presented on LP.........the more I find myself coming up with new ideas of my own. And if you knew what some of those ideas were hahaha  you would probably laugh at me.

It's as though I go thru each day looking at everything in a whole new way......i.e. how various things can be applied to the lottery and if they can help me find a better way to pick numbers.

As for "we of lesser intellectual strength" LOL you CAN'T be talking about yourself. I appreciate your wisdom and the ideas you present. I enjoy reading your posts and will continue to do so.  I am here to learn  and you are one of the better teachers!

Take care and please continue to share your ideas!

in principle it should be possible to codify all different models present on the lottery post by all members, and build complex (but finite) neural network matrix, train it via predictions/results iterations, instruct it to favour best fit and eject non-performers, neural network, if created properly will with certainity arrive with best models based on mathematical odds, such as play pick 3 lottery, or play pairs in pick 4 lottery, because the best performers will fall in those 2 games with certainity. there is a mathematical deep theorem that shows that neural network has certain finite bounds (by incompletness Godel and cardinal numbers and others) beyond which its evolution is limited to the supplied input (in sample case ideas derived from the LP regarding lottery predictions) actually perfectly tuned neural network should after very large training n-iterations approach some basic results from statistics and probability theory (that is re-prove em via numerical means) but no more! there is an intense work done in the area of automatic theorem proving and artificial intelligence in general that utilizes deep ideas from logic and so called category theory and lottery games modeling/simulation could be useful insofar as studying random matrices/stochastic evolution and potentially chaos theory..

As your "401-k" becomes "uh-oh, no-way", you have to expand your horizons a bit.

Here's what Raven62 was aiming at... http://www.lotterypost.com/thread/113989/391541

Go for it. Whatever new way to look at the lottery will benefit your game play.

What types of Expert Systems are there?

1-Bayesian Belief Networks
2-State Machines
3-Backward and Forward: Chaining Inference Engines
4-Case-Based Rules
5-Fuzzy Logic Algorithms
6-Neural Networks

Neural Networks

Neural networks have to do with distributed computing. Instead of having a single processor doing all the work in sequence, you have a bunch of processors doing the work at the same time, and then share the results with the other processors.

Neural networks attempt to mimic the behavior of the brain, processors acting as neurons firing information through synapses.

Most of the work of neural networks is going into pattern recognition and simulations. Neural networks are trained rather than programmed, and potentially will have the most adaptability of any expert system, but like humans they can make mistakes.

Bayesian Belief Networks

This is probably the most difficult expert system to create, and the least common. But that depends on how accurately you define it.

Bayesian Belief Networks are the champions of uncertainty, because they rely on probability rather than definite rules. The expert system doesn't blink if a user can't answer a question: It just goes to the next most likely recommendation without missing a beat.

The knowledge engineers must have a sense for how often given causes occur, and then separately decide how likely they are to know whether or not they occurred.

Haha I remember VRA.

However, I'm more sold on Bayesian Systems. You just need find as many rules, as opposed to numbers, that have greater probability and are more likely to be true.

Oh boy! When I read your post about bee and ant algorithms.......the first thing that popped into my head was getting stung on the nose by a bee and  then the ant movie where the ants are in line doing their thing. A leaf falls in their path and they panic. They all stop and don't know what to do.

Then I thought about all the different ideas that keep popping into my head and how I can apply them to the lottery. LOL here's some examples.....

I was reading a book about tarot cards and the book said sometimes you chose the tarot card, sometimes it chooses you. So I'm thinking hmmm how can tarot cards be applied to the lottery.

I know a little bit about numerology and I started thinking about how the year, month and day of the week could be worked into the previous draw  to come up with some sort of formula.

I have a book with the magic squares and was looking at them wondering what steps could be added to them to come up with numbers.

Last night I was reading a book that mentioned the metaphysical science of numbers. Ok so now I have to find out what that is.

And I am forevermore looking at each days draw for patterns that are present, patterns that are missing, and trying to see whatever it is I don't see.   

Hows that for expanding my horizons? LOL

 I don't believe there's one magical formula to solving the lottery. With so many variables involved I think it's gonna take a few formulas that work together. I also believe that like everything in life there's more than one way to solve this lottery mystery. And I really do appreciate all the ideas that are presented (even if I don't initially understand them) because one idea leads to another. I don't know all the answers and I don't pretend to. I do know that I'm here to learn......... I have learned........ and I continue to learn: )

BTW thank-you for the above link. I took a quick look at it and I'm going to go back and read it more thoroughly.

A long time ago I took a look at the Attrasoft Predictor, but only a quick look, the demo was said to have only "1000 Neurons" my testing showed to me that 1000 neurons were too few for even one position pick 3 prediction, but it looked as if 100, 000 to 1, 000, 000 might be able to help some as you then might get better prediction resolution (Accuracy) and that it might not help for just any kind of prediction and or lottery prediction, but for some and at times, but not always, it might depend on the results given by the program, their interpretation and the technique used for getting them with the Attrasoft Predictor, the results given might tell when to use the results and when not to, the Attrasoft has some flexibility, it is not for sure that very many more "neurons" would help, but the little testing that I did seemed to indicate that there was a good chance that it might.

Can't tell just how good it would be, without first testing the 100 000 or 1 000 000 neurons Attrasoft.