Howdy! I’m a novice programmer and I’m starting to write my very own HTM systems. Recently, I’ve become obsessed with HTM! (I watched HTM school twice in one sitting…) I have a newfound passion for coding, ML and HTM specifically, and an old love for stock market analysis. I’ve decided to start a new hobby where I combine the two! I will begin working on a project where I use HTM systems to predict S&P 500 movements on a day-to-day basis. The purpose of this thread is for me to share my work, ask for assistance, and open a dialogue with the HTM community. I look forward to this journey I’m starting with y’all!
See the bottom of this page.
We used to have an application that did this type of thing. The source code is here.
Yes! Thank you. I’ve seen this briefly before and would like to build off of it further, if at all possible. I’m exploring the possibility for HTM systems to predict S&P 500 prices for the following day; using technical analysis, company fundamentals, market news, time, and other pertinent variables as input. I believe that goal differs from Taurus. I appreciate any arguments and/or input to this idea!
@PhilGoddard Might have some input, I believe he has worked in this arena?
Lots of interesting potential applications of HTM in computational finance.
We’ve had some success in incorporating HTM into trading strategies. However, as with any attempt to make predictions about processes that are inherently noisy(i.e. financial time-series price or returns data), there are many non-trivial issues to overcome and you’ll most likely go down many dead-ends before arriving at anything resembling a profitable strategy.
I always wonder things like: will the effect on the market a clueless leader imposing tariffs be predicted be such a program?
I would guess no but perhaps there is a deeper structure that is not obvious to the casual observer.
@Khoody you had any success on that matter?