Van_der_Voort_4 Maybe the topic serves another purpose and not just the obvious. However we are not selling our algo's but are willing to point those with nobility in the right direction.
Well, then don't post an equity curve if you don't want to discuss anything.what is your point? Post stuff that is open for debate and exchange or don't post. Expect harsh pushback by people when you try to bait them but subsequently state you are actually not out fishing.
GRULSTMRNN My post was more inline of having an intuitive conversation about ML and trading but I see from the majority of negative replies here that this was not the right forum. It seems that most traders here still think the earth is flat. However, for the few who have contacted me wanting to learn about neuromorphic algos, the post was well worth the effort.
That is a strawmen argument. If you wanted to meaningfully discuss ML then why post an equity curve? By any standard, that equity curve looks extremely strange, whatever way you look at it. So far you have not shown at all that you have any value to add to a discussion on ML. Start talking about ML and you will get more reasonable responses. There are a few guys, me included, who work in the space of ML and deep neural networks every single day
GRULSTMRNN Great so are you guys part of the Loihi test environment, or are you heading that way? I get what you are saying about equity curve. Part of our studies has to do with psychological trading, you might note that while the main purpose of the post was to see how many participants would take the vote for manual traders, the vote got the least attention. I posted results of our least profitable bot and I get comments like yours, we have 2 other bots that are more aggressive, but they are not as stable as the first. I must say that this is not our groups only post. Our evaluation has already served its purpose so further replies to this topic may be slow. Just a hint, if you are into ML you should seriously take a look at Loihi this chip offers non binary algos that promises to change trading as we know it. However this will not be self evident until you avail yourself of the dimensional possibilities. Cheers.
I peruse a multi gpu server setup with 4 Titan Rtx and 4 Titan V cards and work mostly with tensorflow and pytorch. I took a glance at Loihi but found it a very immature project. Do you have links to performance tests of real world examples? Are there python apis or language references to look at?
This guy has been talking about raising funds for a hedge fund in previous post and clearly on the DL. Those results do not look real. Panama? Can you say “please scam me”? Dude NO ONE raises money like this except people looking for marks. This has slime written all over it. Good luck with your magic beans, just keep it to yourself.
I emailed him, pretending to be a mark....he gave me a fake name....and a bunch of nonsense. Would give no details before I started a foreign company....lol
I looked into havens a long long time ago. It all seemed to be about tax evasion (or at least secrecy) vs tax avoidance. If that's wrong, or if things have changed, would you be willing to pass along references on how that's done on the up-and-up? I'm not willing to give up a good night's sleep
Thanks for the share. Out of curiosity, how often the model is re-trained? Fellow ML Data Scientist here.