onebornfree
VIP Member
Joined: May 20th, 2013, 8:30 am

January 11th, 2018, 6:15 pm #21

z1235 wrote:
January 11th, 2018, 4:49 pm
I think I get it. There's no point. 
OK. What a shame, I assumed you had a working brain. My bad! Well there goes that theory onto the trash heap of history- to join the " algorithms can predict the future", "Dow theory can predict the future", "head and shoulders patterns can predict the future" , "Kondratieff waves can predict the future", "technical analysis can predict the future" and similar "predictive" investment and speculation theories. :-) .

Regards, onebornfree
Onebornfree's Financial Safety Services: http://onebornfreesfinancialsafetyreports.blogspot.com/

Onebornfree's Special Scam Alerts: http://onebornfree-mythbusters.blogspot.com/

Personal Freedom Consulting: "The Problem Solver: http://onebornfree.blogspot.com/

Onebornfree's 9/11 Research Review :http://onebornfrees911researchreview.blogspot.com/
Quote
Like
Share

Joined: December 12th, 2012, 11:32 pm

January 11th, 2018, 6:37 pm #22

onebornfree wrote:If, for the sake of argument, I assume that your assertion is true, what  does that mean that you  know "for certain" to be  true regarding market behavior in general, in your opinion?
My point is mostly pedantic and has to do with the nature of absolute randomness and how we define choice. That said, human behavior is much more predictable than the market-randomness theorists would dare to imagine. Let's suppose that Bob buys his wife some flowers. Can we figure out why Bob bought her flowers? The most likely reason that Bob bought his wife flowers is that it is their anniversary. It could be due to another reason, of course, but we would be right a non-negligible proportion of the time if we just bet "anniversary". So, already, we see the sketches of algorithmic decision-making. We can formulate these snippets of algorithmic choices as predicates. "Something that people sometimes do for an anniversary is buy flowers for their spouse." This snippet of common-sense is as much a driver of Bob's decision to buy flowers as it is a reflection of a statistical fact about human behavior.
Matthew 5:26
Quote
Like
Share

Joined: February 1st, 2013, 3:01 pm

January 12th, 2018, 1:22 am #23

Wait, why is that common sense? How do we know that we would be right a non-negligible proportion of the time?
Roy: Hello? IT. Have you tried turning it on and off again?
Quote
Like
Share

Joined: December 12th, 2012, 11:32 pm

January 12th, 2018, 2:20 am #24

DanielMuff wrote: Wait, why is that common sense? How do we know that we would be right a non-negligible proportion of the time?
I should have said "common-sense reasoning" - see here. To your second question, perhaps I should be more specific - I know I would be right a non-negligible proportion of the time because I know that giving flowers on a wedding anniversary is a common practice. This is a piece of knowledge about the world that I have by virtue of growing up in this particular region of the world. Other people from other parts of the world wouldn't necessarily be able to make this inference. But the generality of common sense was not what I had in view, merely the fact that common practices (the knowledge of which is part of common sense) drive individual decision-making. It's why you know there will be a lot of people on the roads and malls on Black Friday. You can make this amazing prediction about mass human behavior on the basis of common sense reasoning about a common practice in your culture. The other point is that this is algorithmic. Everybody doesn't just wake up on Black Friday and "feel for no particular reason" like going shopping, as though it were any other day of the year. They go because it's a common practice and the knowledge about this practice is common knowledge.
Matthew 5:26
Quote
Like
Share

onebornfree
VIP Member
Joined: May 20th, 2013, 8:30 am

January 12th, 2018, 6:28 am #25

Clayton wrote:
January 11th, 2018, 6:37 pm
human behavior is much more predictable than the market-randomness theorists would dare to imagine.
Therefor, using "algorithmic thinking", market movements are knowable ahead of time?

Regards, onebornfree
Last edited by onebornfree on January 12th, 2018, 7:46 am, edited 2 times in total.
Onebornfree's Financial Safety Services: http://onebornfreesfinancialsafetyreports.blogspot.com/

Onebornfree's Special Scam Alerts: http://onebornfree-mythbusters.blogspot.com/

Personal Freedom Consulting: "The Problem Solver: http://onebornfree.blogspot.com/

Onebornfree's 9/11 Research Review :http://onebornfrees911researchreview.blogspot.com/
Quote
Like
Share

z1235
VIP Member
z1235
VIP Member
Joined: December 16th, 2012, 3:47 pm

January 12th, 2018, 9:39 am #26

OBF, if you have a reason for doing something, then you're using algorithmic thinking. The only way to avoid algorithms is by doing things for no reason whatsoever. You operate daily, and make decisions just fine by not knowing anything for certain ahead of time. Drop your irrational obsession with certainty already. It's a strawman.
Quote
Like
Share

Joined: December 12th, 2012, 11:32 pm

January 12th, 2018, 11:40 am #27

onebornfree wrote:
Clayton wrote: human behavior is much more predictable than the market-randomness theorists would dare to imagine.
Therefor, using "algorithmic thinking", market movements are knowable ahead of time?

Regards, onebornfree
Knowable in principle, yes. That is, an omniscient being with unlimited contemplation could deduce the future movements of markets. I'll take it one step further - I can imagine an ordinary human organization with sufficient compute resources to model enough of common human behavior to be able to make meaningful predictions about market movements (among other things) before they happen. Such predictions do not need to be long-range to be extremely useful. If the predictions are detailed (and reliable), even knowing a few minutes ahead of time is enough.

Prediction can be recast as a betting-game:

Case A) Without an AI-based market prediction machine, Bob's binary guesses about whether a stock will go up or down are right about 50% of the time and wrong about 50% of the time
Case B) With the aid of an AI-based market prediction machine, Bob's binary guesses about whether a stock will go up or down are right about 50.1% of the time and wrong about 49.9% of the time

In case A, Bob cannot profit long-run by making binary stock-price predictions. In case B, Bob will inevitably profit, as long as he keeps investing in binary markets. This illustrates that predictions do not need to have preternatural insight, they only need to "tip the scale" to enable a consistently better-than-random return on bets.
Matthew 5:26
Quote
Like
Share

onebornfree
VIP Member
Joined: May 20th, 2013, 8:30 am

January 12th, 2018, 12:02 pm #28

z1235 wrote:
January 12th, 2018, 9:39 am
OBF, if you have a reason for doing something, then you're using algorithmic thinking. The only way to avoid algorithms is by doing things for no reason whatsoever.
This statement is incorrect . While so-called algorithmic thinking is indeed one method a person might use to try to problem solve [ either consciously or subconsciously, successfully or unsuccessfully] - heuristic thinking is yet another way that commonly used [consciously or subconsciously, successfully or unsuccessfully], apparently:



Others methods commonly used to try to problem-solve/make predictions include fortune-telling [tea leaves, tarot, I-ching etc.] , and in the area of stock-picking and related, blind-fold dart throwing and similar, or even [horror of horrors !] using a "stock broker or an "investment advisor" :-) .

regards, onebornfree
Last edited by onebornfree on January 12th, 2018, 12:13 pm, edited 1 time in total.
Onebornfree's Financial Safety Services: http://onebornfreesfinancialsafetyreports.blogspot.com/

Onebornfree's Special Scam Alerts: http://onebornfree-mythbusters.blogspot.com/

Personal Freedom Consulting: "The Problem Solver: http://onebornfree.blogspot.com/

Onebornfree's 9/11 Research Review :http://onebornfrees911researchreview.blogspot.com/
Quote
Like
Share

onebornfree
VIP Member
Joined: May 20th, 2013, 8:30 am

January 12th, 2018, 12:05 pm #29

Clayton wrote:
January 12th, 2018, 11:40 am

Prediction can be recast as a betting-game:

Case A) Without an AI-based market prediction machine, Bob's binary guesses about whether a stock will go up or down are right about 50% of the time and wrong about 50% of the time
Case B) With the aid of an AI-based market prediction machine, Bob's binary guesses about whether a stock will go up or down are right about 50.1% of the time and wrong about 49.9% of the time

In case A, Bob cannot profit long-run by making binary stock-price predictions. In case B, Bob will inevitably profit, as long as he keeps investing in binary markets. This illustrates that predictions do not need to have preternatural insight, they only need to "tip the scale" to enable a consistently better-than-random return on bets.
So for yourself, its all about having/finding the "right" "AI-based market prediction machine", correct?

Regards, onebornfree
Onebornfree's Financial Safety Services: http://onebornfreesfinancialsafetyreports.blogspot.com/

Onebornfree's Special Scam Alerts: http://onebornfree-mythbusters.blogspot.com/

Personal Freedom Consulting: "The Problem Solver: http://onebornfree.blogspot.com/

Onebornfree's 9/11 Research Review :http://onebornfrees911researchreview.blogspot.com/
Quote
Like
Share

z1235
VIP Member
z1235
VIP Member
Joined: December 16th, 2012, 3:47 pm

January 12th, 2018, 12:15 pm #30

onebornfree wrote:
z1235 wrote: OBF, if you have a reason for doing something, then you're using algorithmic thinking. The only way to avoid algorithms is by doing things for no reason whatsoever.
This statement is  incorrect . While so-called algorithmic thinking is  indeed one method a person might use to try to problem solve [ either consciously or subconsciously, successfully or unsuccessfully] - heuristic thinking is yet another way that  commonly used [consciously or subconsciously, successfully or unsuccessfully], apparently:

Others methods commonly used to try to problem-solve/make predictions include  fortune-telling [tea leaves, tarot, I-ching etc.] , and in the area of stock-picking and related, blind-fold dart throwing and similar, or even [horror of horrors !] using a "stock broker or an "investment advisor" :-) .
How does this show that my statement is incorrect? 
Quote
Like
Share

z1235
VIP Member
z1235
VIP Member
Joined: December 16th, 2012, 3:47 pm

January 12th, 2018, 12:23 pm #31

Clayton wrote: Knowable in principle, yes. That is, an omniscient being with unlimited contemplation could deduce the future movements of markets. I'll take it one step further - I can imagine an ordinary human organization with sufficient compute resources to model enough of common human behavior to be able to make meaningful predictions about market movements (among other things) before they happen. Such predictions do not need to be long-range to be extremely useful. If the predictions are detailed (and reliable), even knowing a few minutes ahead of time is enough.

Prediction can be recast as a betting-game:

Case A) Without an AI-based market prediction machine, Bob's binary guesses about whether a stock will go up or down are right about 50% of the time and wrong about 50% of the time
Case B) With the aid of an AI-based market prediction machine, Bob's binary guesses about whether a stock will go up or down are right about 50.1% of the time and wrong about 49.9% of the time

In case A, Bob cannot profit long-run by making binary stock-price predictions. In case B, Bob will inevitably profit, as long as he keeps investing in binary markets. This illustrates that predictions do not need to have preternatural insight, they only need to "tip the scale" to enable a consistently better-than-random return on bets.
His name is Jim (Simons), not Bob (though Bob Mercer who works with him is not too shabby either) -- and I don't think he used much "AI" over the years. :)

https://en.wikipedia.org/wiki/Renaissance_Technologies
https://en.wikipedia.org/wiki/Renaissan ... llion_Fund

"Renaissance's flagship Medallion fund, which is run mostly for fund employees,[7] "is famed for one of the best records in investing history, returning more than 35 percent annualized over a 20-year span".[4] From 1994 through mid-2014 it averaged a 71.8% annual return.[8]"

"From 2001 through 2013, the fund’s worst year was a 21 percent gain, after subtracting fees. Medallion reaped a 98.2 percent gain in 2008, the year the Standard & Poor’s 500 Index lost 38.5 percent."

Simons' net worth is close to $20 billion. Probably a dozen more billionaires at RenTech, too. 

Many more like this out there, though not as good.
Quote
Like
Share

onebornfree
VIP Member
Joined: May 20th, 2013, 8:30 am

January 12th, 2018, 12:30 pm #32

z1235 wrote:
January 12th, 2018, 12:15 pm

How does this show that my statement is incorrect? 
Because its wholly exclusive [ of other decision-making methods], not inclusive. If a person has "a reason for doing something", it is not a "sure-fire" certainty that he/she would use "algorithmic thinking" before so doing.
Something generally characterized as "heuristic thinking" by psychologists might also be used, as might other [even more random] means of reaching a decision to act [ listed at the close of previous post].

Regards, onebornfree
Onebornfree's Financial Safety Services: http://onebornfreesfinancialsafetyreports.blogspot.com/

Onebornfree's Special Scam Alerts: http://onebornfree-mythbusters.blogspot.com/

Personal Freedom Consulting: "The Problem Solver: http://onebornfree.blogspot.com/

Onebornfree's 9/11 Research Review :http://onebornfrees911researchreview.blogspot.com/
Quote
Like
Share

Joined: December 12th, 2012, 11:32 pm

January 12th, 2018, 12:34 pm #33

onebornfree wrote:
z1235 wrote: OBF, if you have a reason for doing something, then you're using algorithmic thinking. The only way to avoid algorithms is by doing things for no reason whatsoever.
This statement is  incorrect . While so-called algorithmic thinking is  indeed one method a person might use to try to problem solve [ either consciously or subconsciously, successfully or unsuccessfully] - heuristic thinking is yet another way that  commonly used [consciously or subconsciously, successfully or unsuccessfully], apparently:



Others methods commonly used to try to problem-solve/make predictions include  fortune-telling [tea leaves, tarot, I-ching etc.] , and in the area of stock-picking and related, blind-fold dart throwing and similar, or even [horror of horrors !] using a "stock broker or an "investment advisor" :-) .

regards, onebornfree
So, this post shows that we're talking past each other. The word "algorithmic" in "algorithmic trading" is in contrast to human intuition (or human-based decisions/insight). So, the contrast is between "algorithmic trading" and "human choice-based trading" (on whatever basis the human makes his/her choice). The pedantic point I raised about algorithms has nothing to do with psychology - in this context, "algorithmic" is in contrast to "uncomputable", i.e., problems for which no algorithm can exist (provably). Proving that a sequence of symbols is random, for example, is a non-algorithmic (uncomputable) problem. Proving that two algorithms compute the same function (same result over all inputs) is another uncomputable problem. The space of computable problems, in computer science jargon, is an unimaginably large space that encompasses all possible concepts of order and structure - if a problem is structured (in any way), it is computable because that structure can be exploited by some rule, however difficult it is to derive that rule. Thus, if there is any structure to human decision-making, it is computable... there is some algorithm that describes their decision-making process, even though the problem of finding that algorithm is, in the general case, an uncomputable problem!

No human being can ever provably exhibit an actual instance of an uncomputable problem. To put this concretely, if we locked a person in a room with nothing but a pencil and pad, with instructions to write down a very long sequence of numbers on the pad (say, 1,000 digits), no person could ever write down a number that is absolutely random (algorithmically incompressible) even though almost all numbers of length 1,000 digits are algorithmically random. Even if they did write down such a number, we would never be able to prove it. And what is true for one individual is true for any number of individuals. No matter how many people you employed to do this task at once, the result would still be a compressible (non-random) number. Even if they exchange their notes back and forth and mix them all up. Even if they employed an encryption cipher like AES to encrypt their numbers as they hand them back and forth. Place as many constraints as you like on the process and the result will still be non-random (computable, predictable) from the point-of-view of computability theory. Translating back to the real world, even if we imagined that every market participant were doing nothing but trying to be as unpredictable as possible, the view of computability theory is that their actions are still predictable, individually and in aggregate. Thus, the market is, in principle, predictable.

Heuristics are just guessing rules that hopefully help more than they hurt. We use heuristics in computer algorithms (surprise!), usually to generate "pretty good" solutions to ridiculously hard computational problems. Logic minimization is an example of one field where heuristics are heavily utilized.
Matthew 5:26
Quote
Like
Share

Joined: December 12th, 2012, 11:32 pm

January 12th, 2018, 12:45 pm #34

z1235 wrote: His name is Jim (Simons), not Bob (though Bob Mercer who works with him is not too shabby either) -- and I don't think he used much "AI" over the years. :)

https://en.wikipedia.org/wiki/Renaissance_Technologies
https://en.wikipedia.org/wiki/Renaissan ... llion_Fund

"Renaissance's flagship Medallion fund, which is run mostly for fund employees,[7] "is famed for one of the best records in investing history, returning more than 35 percent annualized over a 20-year span".[4] From 1994 through mid-2014 it averaged a 71.8% annual return.[8]"

"From 2001 through 2013, the fund’s worst year was a 21 percent gain, after subtracting fees. Medallion reaped a 98.2 percent gain in 2008, the year the Standard & Poor’s 500 Index lost 38.5 percent."

Simons' net worth is close to $20 billion. Probably a dozen more billionaires at RenTech, too. 

Many more like this out there, though not as good.
Nice - in my illustration, "Bob" is supposed to be someone who really shouldn't be investing in binary markets... but if we have a good enough AI tool, even our incompetent investor Bob can make money. Of course, then the whole problem of such AI predictors goes back to market dynamics (because it will be profitable for everyone who is an incompetent investor to buy one of these, and their effect will be canceled out, ad nauseum). This is an important thought-experiment when gaming out the future of AI in investing markets. It is easy to see, for example, that AI used by investors will never be open-source because if your competitors have your secret sauce, too, then your secret sauce cannot be profitable. By the same token, we can see that all open-source AI systems will eventually be leveraged by all large-cap investing firms because if it is possible to predict a pattern and you're not doing that, you're leaving money lying on the ground (EMH-style argument). For these reasons, I predict that, at some point, the demand-curve for AI is going to become completely inelastic (price not an object). We've kind of seen this happen with the recent rounds of head-hunting, buying up AI researchers at top-dollar. But I am convinced that this is just a portent of things to come.
Matthew 5:26
Quote
Like
Share

onebornfree
VIP Member
Joined: May 20th, 2013, 8:30 am

January 12th, 2018, 12:51 pm #35

Clayton wrote:
January 12th, 2018, 12:34 pm

So, this post shows that we're talking past each other.
For the record, I'm fully aware of the difference between "algorithmic thinking" in humans and machines. The original post [concerning the use of algorithms for computerized trading systems] that got this whole discussion going I will now re post:

""The Algorithm Scam":

"The discovery of an algorithm that would consistently generate outsized returns would be the holy grail of investing. The individual who came up with such an algorithm and published the methodology in a peer-reviewed journal would be a serious candidate for the Nobel Memorial Prize in Economic Sciences. To date, no one has done so.

This is not surprising. As my colleague Larry Swedroe noted, the market is "forward looking" and incorporates all publicly available information into current prices. The issue is not (as many investors believe) whether news is good or bad. It's whether the news is "better or worse than already expected."

No algorithm can accurately predict surprising news. By definition, a "surprise" is both random and unpredictable...."

"Notwithstanding this compelling logic, investors are often seduced by the "algorithm scam." A broker, advisor or hedge fund operator claims to have a "black box" that will predict future prices. Sometimes the people making these claims have superb credentials, which make their pitch more appealing.........."

"....This type of fraud is particularly insidious because it's difficult to verify. If you ask the creator of an algorithm to tell you how it works, he or she will often claim it's "proprietary" and instead will show you back-tested data supporting its efficacy. The problem with back-tested data is that it produces hypothetical results by showing how investments would have performed if the strategy at issue were implemented. It's easy to misrepresent back-tested data...."
https://www.huffingtonpost.com/dan-soli ... 74582.html

Following this post , you yourself said :

"Every choice you make is the outcome of some algorithm. This is true by definition unless you believe your decisions are somehow channeling uncomputable quantum randomness... " ,

which for myself meant that you believe that all human choices are inevitably also made by an individuals "algorithmic" thinking processes. So thus far, it seems to me that _you_ are the person who mixed/equated the processes of machines with that of humans- but maybe I missed something :-)

Regards, onebornfree
Onebornfree's Financial Safety Services: http://onebornfreesfinancialsafetyreports.blogspot.com/

Onebornfree's Special Scam Alerts: http://onebornfree-mythbusters.blogspot.com/

Personal Freedom Consulting: "The Problem Solver: http://onebornfree.blogspot.com/

Onebornfree's 9/11 Research Review :http://onebornfrees911researchreview.blogspot.com/
Quote
Like
Share

z1235
VIP Member
z1235
VIP Member
Joined: December 16th, 2012, 3:47 pm

January 12th, 2018, 12:52 pm #36

onebornfree wrote:
z1235 wrote:
How does this show that my statement is incorrect? 
Because its wholly exclusive [ of other decision-making methods],  not inclusive.  If a person has "a reason for doing something", it is not a "sure-fire" certainty that he/she would use "algorithmic thinking" before so doing.
Something generally characterized as "heuristic thinking" by psychologists might also be used, as might other [even more random] means of reaching a decision to act [ listed at the  close of previous post].

Regards, onebornfree
Algorithms can (and do) use heuristics. Actually, plenty of optimization algorithms use heuristics when a closed form mathematical solution does not exist or is difficult to find. No one in mathematics or computer science would describe/define algorithms as smth "wholly exclusive" with heuristics. 

Every time you use an if/then logic in your actions/decisions you're using an algorithm, even when "if" = "feels like it's a good idea". Do you think that your Permanent Portfolio does NOT use an algorithm, and that it makes no assumptions about the future? 
Quote
Like
Share

Joined: December 12th, 2012, 11:32 pm

January 12th, 2018, 12:53 pm #37

onebornfree wrote: So for yourself, its all about having/finding the "right" "AI-based market prediction machine", correct?
I'm not an investor. I learned a long time ago that investing just isn't something I'm cut out to do. But I love technology and I am endlessly fascinated by the idea of leveraging information theory to understand the world at deeper and ever deeper levels. I expect that within my lifetime, machine-guided human intelligence will be so far superior to seat-of-the-pants thinking that no one will ever use the latter for commercial purposes*. It would be like trying to play chess against somebody who is using the Alpha Zero chess-engine to help them choose their next move.

*The exception is for those services where "the human touch" can make the difference, but even these are going to get "Starbucked" ... :(
Matthew 5:26
Quote
Like
Share

Joined: December 12th, 2012, 11:32 pm

January 12th, 2018, 1:02 pm #38

Related to my last post:

Matthew 5:26
Quote
Like
Share

z1235
VIP Member
z1235
VIP Member
Joined: December 16th, 2012, 3:47 pm

January 12th, 2018, 1:13 pm #39

Clayton wrote:
z1235 wrote: His name is Jim (Simons), not Bob (though Bob Mercer who works with him is not too shabby either) -- and I don't think he used much "AI" over the years. :)

https://en.wikipedia.org/wiki/Renaissance_Technologies
https://en.wikipedia.org/wiki/Renaissan ... llion_Fund

"Renaissance's flagship Medallion fund, which is run mostly for fund employees,[7] "is famed for one of the best records in investing history, returning more than 35 percent annualized over a 20-year span".[4] From 1994 through mid-2014 it averaged a 71.8% annual return.[8]"

"From 2001 through 2013, the fund’s worst year was a 21 percent gain, after subtracting fees. Medallion reaped a 98.2 percent gain in 2008, the year the Standard & Poor’s 500 Index lost 38.5 percent."

Simons' net worth is close to $20 billion. Probably a dozen more billionaires at RenTech, too. 

Many more like this out there, though not as good.
Nice - in my illustration, "Bob" is supposed to be someone who really shouldn't be investing in binary markets... but if we have a good enough AI tool, even our incompetent investor Bob can make money. Of course, then the whole problem of such AI predictors goes back to market dynamics (because it will be profitable for everyone who is an incompetent investor to buy one of these, and their effect will be canceled out, ad nauseum). This is an important thought-experiment when gaming out the future of AI in investing markets. It is easy to see, for example, that AI used by investors will never be open-source because if your competitors have your secret sauce, too, then your secret sauce cannot be profitable. By the same token, we can see that all open-source AI systems will eventually be leveraged by all large-cap investing firms because if it is possible to predict a pattern and you're not doing that, you're leaving money lying on the ground (EMH-style argument). For these reasons, I predict that, at some point, the demand-curve for AI is going to become completely inelastic (price not an object). We've kind of seen this happen with the recent rounds of head-hunting, buying up AI researchers at top-dollar. But I am convinced that this is just a portent of things to come.
Surprisingly (to many, but not me) AI hasn't (yet) lived up to its holy grail expectations for trading/investing. Here's an example (sorry for the bad copy/paste but the article behind a WSJ paywall so a link wouldn't have worked):

The Future Is Bumpy: High-Tech Hedge Fund Hits Limits of Robot Stock Picking
Voleon is among investors deploying machine learning, a technology in which computers develop trading strategies. It’s harder than it sounds.

By  
Bradley Hope and 
   
Juliet Chung
Updated Dec. 11, 2017 7:23 p.m. ET 
52 COMMENTS

BERKELEY, Calif.—For Michael Kharitonov, building a hedge fund based on machine learning has been a rule of threes: It was three times as hard, and it took three times as long, as anticipated.
“Most of the things we’ve tried have failed,” said the co-founder of a little-known firm called Voleon Group.
Machine learning, a set of techniques that empowers computers to find patterns in data without using rules prescribed by humans, has been producing advances in a range of fields, from robotics to weather forecasting to language translation. The technique is at the heart of efforts to build self-driving cars. 
Why not use it to crack financial markets? The notion has led to an arms race of sorts, as multibillion-dollar investment firms that already were mathematically focused have been signing up the smartest computer scientists and statisticians they can find.
The gambit seems to be working for two of this year’s top-performing hedge funds. Quantitative Investment Management LLC, up 68% this year in its biggest fund, attributes its success to the technique. Teza Capital Management LLC credits machine learning in part for its more than 50% gain so far this year.
<... snipped for fair-use ...>

This “doesn’t mean we don’t think about what’s going on,” Mr. McAuliffe said. Voleon researchers sometimes design what they call “perturbations” to study the importance of various inputs into the prediction system.
This testing also helps them figure out whether, in certain cases, it might be too tuned to historical data to be useful for forecasts. In statistics, this is known as the problem of “overfitting.”
Voleon’s computers look for relationships in not just financial information but also nonfinancial data sets. Broadly, these could include anything from analyses of satellite images and shipping manifests to credit-card receipts and social-media sentiment about particular companies. Successfully analyzing such data is a goal of quants and non-quants alike as they look for a heads-up about changes in the health of an industry or the supply of a commodity. 
The Voleon principals won’t reveal what data they feed into their system, to say nothing of how they have trained it to assess the data. Like other quant firms, Voleon guards its techniques and strategies. Frosted glass on its quarters provides privacy. No sign on the property identifies the firm. 
Investors uncomfortable with the mystery of it all have “self-selected” out of the firm, Mr. Kharitonov said. While he understands their discomfort, he believes computers make fewer mistakes than people. 
“The application of machine learning science to financial prediction is still in its early stage,” he said. “We are just scratching the surface.” 
Last edited by z1235 on January 12th, 2018, 1:59 pm, edited 2 times in total.
Quote
Like
Share

onebornfree
VIP Member
Joined: May 20th, 2013, 8:30 am

January 12th, 2018, 1:36 pm #40

z1235 wrote:
January 12th, 2018, 12:52 pm
onebornfree wrote:
z1235 wrote:
How does this show that my statement is incorrect? 
Because its wholly exclusive [ of other decision-making methods],  not inclusive.  If a person has "a reason for doing something", it is not a "sure-fire" certainty that he/she would use "algorithmic thinking" before so doing.
Something generally characterized as "heuristic thinking" by psychologists might also be used, as might other [even more random] means of reaching a decision to act [ listed at the  close of previous post].

Regards, onebornfree
Algorithms can (and do) use heuristics. Actually, plenty of optimization algorithms use heuristics when a closed form mathematical solution does not exist or is difficult to find. No one in mathematics or computer science would describe/define algorithms as smth "wholly exclusive" with heuristics. 

Every time you use an if/then logic in your actions/decisions you're using an algorithm, even when "if" = "feels like it's a good idea". Do you think that your Permanent Portfolio does NOT use an algorithm, and that it makes no assumptions about the future? 
So psychologists differentiate between the two ways of decision making for humans for no good reason, correct? There are no distinctions- for humans, heuristic decision making is automatically incorporated into all "algorithmic" thought processes, as you see it, right?

And, algorithmic thinking for you also includes the "it feels like a good idea" notion, right?

regards onebornfree
Last edited by onebornfree on January 12th, 2018, 1:41 pm, edited 1 time in total.
Onebornfree's Financial Safety Services: http://onebornfreesfinancialsafetyreports.blogspot.com/

Onebornfree's Special Scam Alerts: http://onebornfree-mythbusters.blogspot.com/

Personal Freedom Consulting: "The Problem Solver: http://onebornfree.blogspot.com/

Onebornfree's 9/11 Research Review :http://onebornfrees911researchreview.blogspot.com/
Quote
Like
Share