Forecasting currency exchange rates using a neural network - page 2

 
alsu:

"The result is visible... but needs more research" is quite possibly self-defeating. Try to think of yourself as a rudimentary soulless one-liner with two inputs: "how long have I got left in this world" and "the delta cluster has brought me so much profit", and one output: "is it worth wasting time".

If an idea does not immediately yield results, you should drop it and move on to something else. There is no need to regret, if there was a rational basis for it, and there really was "not enough research", the old idea will come back to you sooner or later.


Actually there is no time, and to check for STABILITY of approach, then everything is simple, there are a lot of signs by which it is possible to find out a chance or regularity....The result is received, but the work of a network looks like a chance. It's another matter to find the input so that randomness becomes a pattern....
 
nikelodeon:

In fact, there is no time, but to check for STABILITY of approach, it is simple, there are a lot of signs to know if it is random or regularity.... At the moment the result is obtained, but the work of the network looks like randomness. It's another matter to find the input so that randomness becomes a pattern....

Oh, then it's time to visit us in the next thread, there's a fresh piece about regularities))
 
nikelodeon:



I just want to say that "good thinking comes afterwards": paradoxically, it is often useful to quit in order to achieve results)
 
KID: I don't think you'll find something like that in there, as he made up the subject himself, but I'll look for it.
Maybe something more original to come up with? for the topic of course forecast in 2013 cannot be invented by the professor himself for it is as old as this world and has been discussed by more than one generation (already a generation) of original teachers..... )))
 
nikelodeon: I don't know.... What about the delta cluster? Leonid, have you tried it? I'm trying to build a network right now, I think we're getting a good result... but I need more research...


my friend, where are you going? we've been discussing this topic for years. isn't it clear yet?

Well speaking of the delta cluster, it's not a course prediction in the literal sense of the word anyway ))))

 
KID: This subject has a very specific teacher, which makes the coursework "off the charts", so no one cares about relevance.
Then, your term paper should consist of one sentence: "Rate prediction using a neural network is not possible". ))) That's the whole coursework ))))
 
LeoV: Then, your term paper should consist of one sentence: "Rate prediction using a neural network is not possible". ))) That's the whole coursework ))))

Well why, it's always possible. Another thing is whether there will be any meaningful relationship between input and output...
 
LeoV:
Maybe something more original to come up with? for the topic of course predictions in 2013 cannot be made up by the lecturer himself because it is as old as this world and has been discussed by more than one generation (already a generation) of original lecturers..... )))


The course is not from 2013, but from 2010, in order to be able to check the results with what was in fact the case
 
LeoV:
Then, your coursework should consist of one sentence:"Forecasting the rate with a neural network is not possible". ))) That's the whole coursework ))))

This is the first time I read something like this from LeoV. I even feel a special respect.
 
DhP: This is the first time I've read something like this from LeoV. I even have a special respect for it.

)))

I've never been a fan of predicting financial markets )))) even with the help of neural networks.... )))