Search

Approximating chaotic systemsReply
I had found that time series over a mostly uniform distribution tends to return an average when trying to make a prediction. Had hopes for SARIMAX when studying python time series analysis, but the predictions always came out between 4.8 and 5.2 (for pick 3). Then when you see how predictions based on ARIMA are calculated, you realize it was the wrong tool to use My focus is shifting to classification, though I am not exactly sure what it will look like just yet.
Aug 25, 2023, 8:46 pm - hypersoniq - Mathematics Forum

Megaball predictionReply
Assuming all megaballs are equally likely to be chosen, 1/25 = p = probability that a draw prediction is correct due purely to chance. Let Y = number of correct predictions in n=322 independent draws. The probability distribution of Y = P(Y=y) is binomial with parameters (p, n = 322). Since n is large, P(y = Y) = cumulative distribution function of Y = cdf of Y is very well approximated by the Central Limit Theorem as P(y = Y) = N[(y - n*p) / sqrt(n*p*(1 - p))], where N(x) is the cumulat
May 9, 2023, 1:56 am - Wavepack - Mathematics Forum

Neural Net Lottery PickerReply
Considering that 90% of the data coming from the lottery site is normalized. All predictions all effectively basis against said data. This in turn might get you 2 or 3 numbers maxed.
Jun 25, 2022, 11:02 am - Cyberplex - Mathematics Forum

Neural Net Lottery PickerReply
Could you please describe the type of neural net used, the parameters (#layers, #neurons, training set data size, etc), what lotto you are predicting, and how to interpret your array of numbers that you post? Have you looked at the neural net internals to get an idea of what statistics the NN ends up using for its predictions?
Jan 28, 2023, 6:08 pm - Wavepack - Mathematics Forum

Megaball predictionReply
About a year ago, I did the Chi-square test on Mega Millions and Powerball, both white and non-white balls. I was going from memory that each had too high probability estimates for a non-uniform pdf test. Those test results from a year ago is what motivated me to see if there was any pdf non-uniformity that could be used to improve predictions. It's been a fun side project every since that I work on every now and then. The project drives home probability and statistics fundamentals, along w
May 10, 2023, 5:55 pm - Wavepack - Mathematics Forum

Megaball predictionReply
As stated in the initial post, the predictions were made using only past draw data, and there was no training of the method with data at all (other than adjusting two time window length parameters) Maybe 2 to 3 percent prediction accuracy over expected was due to chance, but unlikely 4% over expected. I've got good statistical significance. Over three sigma. If you do a Chi-square analysis on past Megaball draw data to test the uniform pdf hypothesis, the results suggest there are good o
May 10, 2023, 11:51 am - Wavepack - Mathematics Forum

Chi-Squared Goodness of Fit Test for LotteriesReply
I also implemented 3rd order Bayesean optimal estimation, despite the limited data set. I also implemented a blend between 1st and 2nd order statistics. I also tried joint Bayesean optimal estimation of intra-draw pairs and inter-draw pairs, which you could view as a subset of full 3rd order statistics. I have developed a program that picks in a way to maximize 3W, 4W, and 5W hits since that's what pays. My analysis shows that there exists a tiny bit of information in past draw data
Feb 7, 2023, 2:58 pm - Wavepack - Mathematics Forum