Forums  > Trading  > You Know Who - Renaissance Watch  
     
Page 1 of 41Goto to page: 1, 2, 3 ... 40, 41 Next
Display using:  

gaj


Total Posts: 60
Joined: Apr 2018
 
Posted: 2019-12-18 11:49
One technique that I've often heard associated with RenTech is Hidden Markov Model. Lenny Baum invented Baum-Welch Algorithm which solves the MLE for HMM. Mercer and Brown from IBM worked on speech recognition problems that also use HMM.

Anyone has insights on how HMM is used in trading? What are examples where HMM is superior than simple regression?

rod


Total Posts: 395
Joined: Nov 2006
 
Posted: 2019-12-18 13:33
From another Talking Machines podcast, this one with Nick Patterson, that was previously posted on this thread:

"I joined a hedge fund, Renaissance Technologies. I'll make a comment about that. It's funny that I think the most important thing to do on data analysis is to do the simple things right. So, here's a kind of non-secret about what we did at Renaissance: in my opinion, our most important statistical tool was simple regression with one target and one independent variable. It's the simplest statistical model you can imagine. Any reasonably smart high school student could do it. Now we have some of the smartest people around working in our hedge fund. We have string theorists we recruited from Harvard, and they're doing simple regression. Is this stupid and pointless? Should we be hiring stupider people and paying them less? And the answer is no. And the reason is nobody tells you what the variables you should be regressing [are]. What's the target? Should you do a nonlinear transform before you regress? What's the source? Should you clean your data? Do you notice when your results are obviously rubbish? And so on. And the smarter you are the less likely you are to make a stupid mistake. And that's why I think you often need smart people who appear to be doing something technically very easy, but actually usually it's not so easy."




rod


Total Posts: 395
Joined: Nov 2006
 
Posted: 2019-12-18 13:40
@EspressoLover, are you sure Eric Lander was at RenTech? Are you confusing him with Nick Patterson?

EspressoLover


Total Posts: 391
Joined: Jan 2015
 
Posted: 2019-12-18 15:22
@rod. Thanks for pointing that out. You're right.

Good questions outrank easy answers. -Paul Samuelson

jslade


Total Posts: 1200
Joined: Feb 2007
 
Posted: 2019-12-19 17:26
> C'mon guys. Information geometry? I'm sure there's maybe one small application there using information geometry for some corner case that gives a 2% improvement over regular PCA.

FWIIW there are lots of uses of Information Geometry where you get fiddley and general improvements in bias variance on estimators in the presence of non-normal noise, which is obviously relevant to this sort of thing. For example: click here. I buy that it gets used somewhere, and was glad to hear about it here as it's a cool basket of ideas for my own efforts.

Also, yes, I originally posted the Patterson thingee in this thread, and know people at Edgestream who say the same things. The moon math isn't the important thing, but I'd be shocked and amazed if they didn't work the latest results in here and there.

"Learning, n. The kind of ignorance distinguishing the studious."

nikol


Total Posts: 909
Joined: Jun 2005
 
Posted: 2019-12-19 23:03
Surprised this is not here

https://www.youtube.com/watch?v=SVdTF4_QrTM

His humor sounds as Tom Waits performance at "Nighthawks At the Diner" or it is Tom's sound like a humor of Jim.
Previous Thread :: Next Thread 
Page 1 of 41Goto to page: 1, 2, 3 ... 40, 41 Next