Marek Ka @Rekcross

ricknekbrazil.blogabet.com Joined January 2018

Tweets

295
Followers

44
Following

75
Likes

216

Marek Ka @Rekcross

6 years ago

@12Xpert Hi, I noticed that on your web you have this neat graph. Can you point me to the article which explains analogy behind this? I count find it, thanks!

1 0 0 0 0

View Details

Marek Ka @Rekcross

6 years ago

@BettingIsCool @professor042 I got same answer in excel and with minimize, Im sure that you objective fun is more complex but better to try simple things first and then move to task at hand :)

1 0 0 0 0

View Details

@BettingIsCool @professor042 One more Q why do you have multiple bounds? If you have objective fun with one variable (stake) one min,max pair should be enough. I assume that objective fun is return profit of all bets with given stake

2 0 0 0 0

View Details

Marek Ka @Rekcross

6 years ago

@BettingIsCool @professor042 Okey bounds need to be defined as sequence of min,max pairs :)

1 0 1 0 0

View Details

Marek Ka @Rekcross

6 years ago

@BettingIsCool @professor042 Regarding methods i find scipy documentarion very useful you should be able to pick one for your task, my personal preferences are bfgs and newton optimizations. You mean bounadries [0, inf)? In scipy it should be [0, None], none is for no restrictions on upper boundary

1 0 0 0 0

View Details

Marek Ka @Rekcross

6 years ago

@BettingIsCool @professor042 I guess there is no gdr beside if you code tour own but you can use scipy.optimize.minimize to negative of your function. Like scipy.optimize.minimize(lambda x: -f(x), bounds=[], method=“”)

1 0 0 0 0

View Details

Marek Ka @Rekcross

7 years ago

Picture of @stylebender entrance, the most epic in @ufc history "painted" by AI. What do you think? :) #artificialIntelligence #art #painting

0 0 2 0 0

View Details

Marek Ka @Rekcross

7 years ago

Hmm, public question, how many clusters would you make? 😅4, 5, 7 looks reasonable

0 0 0 0 0

View Details

Marek Ka @Rekcross

7 years ago

@Mo_Nbg @SmarterSig @BettingIsCool well i doubt many people possess dataset that i have ;) at least I havent seen anyone, its kind specific so im not that worried

0 0 0 0 0

View Details

Marek Ka @Rekcross

7 years ago

@SmarterSig @Mo_Nbg @BettingIsCool Yea thats right anyway i see interest in this topic so i will try to post some news from development of my model in summer in case anyone is interested

2 0 2 0 0

View Details

Marek Ka @Rekcross

7 years ago

@Mo_Nbg @BettingIsCool To elaborate my model makes 16% yield on the test set, 750 examples, average odds 1,47. Yea i know it isnt that great 40k history but nevertheless. In deployment its around 9% on 234bets

1 0 0 0 0

View Details

Marek Ka @Rekcross

7 years ago

@Mo_Nbg @BettingIsCool Yea i maybe gone too far with that. What i wanted to express was that its better than woc. That was more or less my point

0 0 0 0 0

View Details

Marek Ka @Rekcross

7 years ago

@Mo_Nbg @BettingIsCool Totally agree Giod quote: no model is perfect but some are useful

0 0 1 0 0

View Details

Marek Ka @Rekcross

7 years ago

@Mo_Nbg @BettingIsCool yeah i guess that my profits are only fluke then :D sure they are only another way of modeling but far more superior than everything i saw produced by joseph, betting is cool or other tipsters. With use of dnns and some error analysis you can make nice money ;)

2 0 0 0 0

View Details

Marek Ka @Rekcross

7 years ago

@KautoStar2007 @BettingIsCool Definitely python and take some classes on probability and statistics. You can find whole courses at Youtube from mit etc. then learn machine learning then deep learning ;) simple models wont make profit at pinnacle

0 0 2 0 0

View Details

Marek Ka @Rekcross

7 years ago

@PinnacleSports 1:3 (Arsenal win), Aubameyang

0 0 0 0 0

View Details

Pinnacle @Pinnacle

7 years ago

"The best handicappers in the business are typically capable of about 57% win rates. After the bookmaker’s margin, that translates into about a 110% return on investment." bit.ly/2zdMWRh

0 6 24 0 0

View Details

Marek Ka @Rekcross

7 years ago

@drvox its nice but what will you do with that?Renewable sources of energy have far too low output and basicly most of them cannt be regulated.So it leaves you with two possibilities:1)decrease power load(stop economic growth) 2)invent thermonuclear fusion/ something similar