Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Using a Keras Long Short-Term Memory Model to Predict Stock Prices

heartbeat.fritz.ai

Post reply on HN
Print view View original

121–130 of 178 posts

Re: Using a Keras Long Short-Term Memory Model to Predict Stock Prices

#121

Post by jdmichal »

jwiley wrote: post #102
goodside wrote: post #93

This is pure nonsense. This isn’t even the right way to begin thinking about this as a forecasting task — the target series should be log-normal returns, not raw asset price. The performance of this model is laughably bad, which is probably why he spends zero time evaluating its effectiveness. You could trivially get better forecasts than this by naively repeating the last-observed price . This isn’t ML. It’s cargo-c…

Can you clarify why "the target series should be log-normal returns" is important or provide a pointer for more information?

Well, just at a minimum... There's the fact that the majority of large gains and losses exist over single-day frames. That is, guessing "right" or "wrong" on movement means little when one slip on the wrong day will decimate your returns.

Re: Using a Keras Long Short-Term Memory Model to Predict Stock Prices

#122

Post by tomazio »

pcprincipal wrote: post #88
DevX101 wrote: post #26

I'm not a purist believer in the efficient market hypothesis. BUT, I doubt there's much alpha to be gained simply from looking at price data which is widely and publicly available. Also keep in mind that markets are dynamic feedback loops so even if this model had an edge, the act of publishing this article would work against you to neutralize that edge in the future. There's a good reason the most successful 'quanti…

I 100% agree with this as a Daily Fantasy Sports player who has published some blog posts and code related to strategy. Any success I've had has been in sports and formats that are not popular and I purposely do not write or open-source code on. Edge can go away almost immediately; I saw this firsthand in DFS when a number of websites came in with free tools that pros had been using for years. I play in the stock mar…

I'm very interested to read some of your blog posts related to strategy/code around DFS, but I don't know what your blog is. Would you mind posting a link to one of these posts?

Re: Using a Keras Long Short-Term Memory Model to Predict Stock Prices

#123

Post by gonehome »

fredliu wrote: post #71
rl3 wrote: post #59

Earlier quoted context omitted.

> Alpha Go (Zero) is very specialized for the game of go, not sure how much of its specialized algorithm could transfer to other generic use cases. AlphaZero is a generalized successor, and it does just that: https://en.m.wikipedia.org/wiki/AlphaZero

Good to know it generalized, but seems it's only generalized on board games like problems where you have a problem space to search through, is that understanding correct? If so, it probably won't help in the use cases we are talking about here, right?

Yeah it's not an AGI, but I think the response was more that DL isn't necessarily limited to automating tasks that are easy for humans but hard for machines without DL (which I think was your original statement).

DL could potentially do a better job by recognizing patterns in the dataset that lead towards winning (making more money) that humans might miss. Like how AlphaZero can recognize moves in Go or Chess that humans don't understand are the best moves to make.

It's not obvious to me how this would be done, but I think it's plausible that some clever implementation could help.

Re: Using a Keras Long Short-Term Memory Model to Predict Stock Prices

#125

Post by qaq »

WrtCdEvrydy wrote: post #115
JumpCrisscross wrote: post #111

Earlier quoted context omitted.

They’re totally up front about this requirement, too, when making offers. (Source: I turned one down.) You are to move to Long Island, buy a house, buy into the employees-only fund and work there for the rest of your financial career.

LOL, No. How do you even enforce that? If I get a job somewhere else, there's no reason for me to disclose where I'm going.

They will prob run regular checks and you are prob. signing off on them having a right to do so.

Re: Using a Keras Long Short-Term Memory Model to Predict Stock Prices

#126

Post by smaddox »

DevX101 wrote: post #53
C1sc0cat wrote: post #50

Earlier quoted context omitted.

A lifetime noncompete really I assume they pay you for life for this noncompete.

My understanding is that legally, the contract says something reasonable, like a year. But the firm will use it's considerable resources to impede your career if you attempt to work in mathematical trading after that year's up. On the flip side, employees are compensated very well, and the turnover is very low.

If that's true, it's only a matter of time until they get sued and lose a large class action suit.

Re: Using a Keras Long Short-Term Memory Model to Predict Stock Prices

#128

Post by sseveran »

jwiley wrote: post #102
goodside wrote: post #93

This is pure nonsense. This isn’t even the right way to begin thinking about this as a forecasting task — the target series should be log-normal returns, not raw asset price. The performance of this model is laughably bad, which is probably why he spends zero time evaluating its effectiveness. You could trivially get better forecasts than this by naively repeating the last-observed price . This isn’t ML. It’s cargo-c…

Can you clarify why "the target series should be log-normal returns" is important or provide a pointer for more information?

Here you go:

https://financetrain.com/why-lognormal-distribution-is-used-...

Re: Using a Keras Long Short-Term Memory Model to Predict Stock Prices

#129

Post by ak39 »

jpeterson wrote: post #114
jwiley wrote: post #102

Earlier quoted context omitted.

Can you clarify why "the target series should be log-normal returns" is important or provide a pointer for more information?

Investment markets operate on relative gain, not absolute gain. E.g.: if you invest in a stock and it gains $5, this would be a great return for a $1 stock but a poor one for a $1000 stock, so the absolute gain doesn't mean anything on its own. A 5% return always means that you've gained 5% on your investment.

Would it be the same (valid) with percentage returns?

Re: Using a Keras Long Short-Term Memory Model to Predict Stock Prices

#130

Post by sseveran »

turingcompeteme wrote: post #92
SEJeff wrote: post #75

Earlier quoted context omitted.

Most of them aren't hedge funds, they're private (HFT, Electronic Trading, Botique Firms, Proprietary Trading are all terms you'll hear) so they don't have public data. It is literally impossible to prove without insider information. However the Virtu Financials, the Citadels, etc, will always exist, and will be doing exceptionally well whether people realize it or not.

I realize this doesn't actually matter and I feel weird defending Renaissance... but Virtu is a public company and hasn't been doing that well, and returns from Citadel's various funds are not hard to find. Also, I don't know how to compare a fund's returns to that of a private business. But Medallion has been around since before HFT was really a thing, and I'm not aware of any HFT places that have been growing at 70…

These are wildly different businesses. Virtu is a global electronic market maker which in reality is more like a mature technology company than a risk taking hedge fund. Renaissance (Medallion specifically) is a weird fund that manages employee money. Citadel is a large, diversified financial services company that happens to have overlap with both of the previously mentioned firms. They have a large market making operation and offer a variety of different hedge funds that they market
Post reply on HN