Econometrics: one step ahead forecast - page 68
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
My criteria has nothing to do with feet at all.
Clearly not, but you have formulated the criteria for a good model. For a model that implies the presence of a stop, those criteria are useless. Or these criteria are only for models which imply prediction for a fixed time. I.e. limited only by time and not by price?
But in order to buy the muv back, you must also split the buyback into 13 iterations. And only the market knows at what level your final redemption muve will be after 13 bars:
Derivatives will ruin the world. (with, me and W. Buffett)
I don't use neural networks and don't believe in them at all.
And you use a linear one-layer forward propagation network with Hodrick-Prescott as input and prediction as output
EURUSD = C(1)*EURUSD_HP(1) + C(2)*D(EURUSD_HP(1)) + C(3)*D(EURUSD_HP(2))
https://en.wikipedia.org/wiki/Feedforward_neural_network
So your statement should read as follows: "I don't believe in non-linear multilayer networks".
clearly not, but you have formulated criteria for the goodness of the model. For a model implying the presence of a stop, these criteria are useless. Or these criteria are only for models that imply prediction for a fixed time. I.e. limited only by time, not by price?
The model's linearity was originally formulated.
... well - so ... in the sense of linearity of the coefficients under the standard functions of the PP package.
Quasi-linearity ...
And you use a linear one-layer forward propagation network with Hodrick-Prescott as input and prediction as output
EURUSD = C(1)*EURUSD_HP(1) + C(2)*D(EURUSD_HP(1)) + C(3)*D(EURUSD_HP(2))
https://en.wikipedia.org/wiki/Feedforward_neural_network
So your statement should read as follows: "I don't believe in non-linear multilayer networks".
No, it's something like this: "I don't believe in non-linear and multi-layered networks". It's hard enough for Greece, at least appreciate the classics :)
No, it's something like this: "I don't believe in non-linear and multi-layered networks". It's hard enough for Greece, at least appreciate the classics :)
You may be right.
It is tantamount to saying that red is better than blue.
And it's more a question of the appropriateness/appropriateness of applying one or the other.
Fuzzy sets solve part of the problem and you have to compare the nut with the auto, not red and blue.
Read the thread from the beginning and the articles on which the thread is based.
No, it's something like this: "I don't believe in non-linear and multilayer networks".
Linear multilayer networks do not exist because they are mathematically reducible to single layer networks. So faa1947's statement should read like this: "I don't believe in non-linear networks". Pretty narrow-minded.
Clearly not, but you have formulated criteria for the goodness of the model. For a model which implies the presence of a stop, these criteria are useless. Or these criteria are only for models which imply prediction for a fixed time. I.e. limited only by time and not by price?
Exactly in line with the theme: predicting one step ahead.
I'm trying to find such characteristics of a model on the history that may say something about the predictability of the model. So far I have given my list. I have no evidence that these properties of the models have anything to do with its predictability, but no evidence to the contrary either. One thing that is clear so far is that it is good if the model has the listed properties.
No, it's more or less like this: "I don't believe in non-linear and multi-layered networks". Greece is hard enough as it is, at least appreciate its classics :)
The distrust of neural networks, we must assume, is due to the difficulty of making this forex dish.
Nets turn out to be understudied.
But, if you don't aim to create a highly intelligent EA, they can turn out very well.
One-cells are not highly intelligent, but they can/can devour some, and their sense of self-preservation helps them evade others.
Perhaps that's what the forex is all about: being able to devour/capture your own in time and escaping without being devoured in time.
Simple forever...