Welcome Guest
( Log In | Register )
The time is now 11:33 am
You last visited January 18, 2017, 9:04 am
All times shown are
Eastern Time (GMT-5:00)

How to teach a supercomputer to read the minds of traders or investors


Last Edited: May 30, 2008, 3:45 am


I'll like for you all to listen to this comment in this youtube.com video by a trader about the holy grail of the Stock market...


How can you teach a supercomputer to read the minds of traders/investors....One good way to do this is to feed the supercomputer TV News data from every part of the world and let the supercomputer filter 1 specific keyword...

But this are not any TV news data, this are news data with synonyms and let the supercomputer translate every data news from different parts of the world into synonyms and if one synonims matches the list of synonims that the supercomputer has in its memory or core, then the supercomputer can know that is bad or good, that is a bad news or good news or ranges from the different emotions...But you see the news data coming from each country in the world has to be translated by a program into synonyms and summarized the synonyms...The news data must be comming in English but still it has to be stripped away from adjective or it can be summarized but you must tell the program to add 1 word and 1 word together and find out if is a good emotion or a bad emotion and then feed it to the supercomputer....You then teach the supercomputer the weight of emotions, wether certain emotions carry a more heavy weight than others and you can rank the emotions....This is just the first step...I know it an be done....Yes it can be done....Years ago i heard of a supercomputer that could read any passage from any book and tell which author it was....How it did it? Because different authors had different style of writing...So if that can be done, i know teaching a supercomputer the difference between emotions from a news data can be achieved...Not only that but you can let the supercomputer, once you thaught it to differentiate the different emotions based on what i read/analyzed from the program feeding it, you can let the supercomputer look at past history of stocks and compare them with the emotion of the day and the stock responses and analyze the reaction of traders/investors and anylize if the reaction was a reaction of fear, caution or of a green light...

I know it can be done...And this is an idea where scientist can elaborate and take it a step further...These are the first step....At first the supercomputer will not be sharp but after the first couple of years it will be on autopilot....

You teach the supercomputer to have this as a back up, as a constraint and to exercise it after it has analyzed what stocks are next for the following day, weeks or months wheather this analyzation is by some formula or previous stock history analyzation..And each day the supercomputer will be fed processed News from the program, from around the world or from America, all sorts of news, bussiness news, technology news, etc....

Entry #165


MillionsWantedComment by MillionsWanted - May 30, 2008, 6:01 am
They probably used an artificial neural network to learn the writing styles of the different authors.

I am experimenting with a neural network addon in MS Excel to see if it can find patterns in the draws.
Usually it is able to find 1 or 2 correct numbers in the draw. This can be useful when playing with systems using groups. If you put the predicted numbers in a group of its own and you know you're almost guaranteed 1 or 2 numbers correct you have eliminated a bunch of lines already.
Comment by pumpi76 - May 31, 2008, 1:45 am
i meant to say to filter for 1 keyword of each documment submitted to it and whether this keyword is an emotion...
Comment by pumpi76 - May 31, 2008, 1:49 am
i meant to say the supercomputer to filter for 1 keyword of each documment submitted to it and whether this keyword is an emotion...
Comment by pumpi76 - May 31, 2008, 10:35 pm
i learn this process is similar to an area called: Data Mining...So it will be similar to data mining....

You must be a Lottery Post member to post comments to a Blog.

Register for a FREE membership, or if you're already a member please Log In.