Machine learning in trading: theory, models, practice and algo-trading - page 3003

[Slava]

Stanislav Korotky #:
Please clarify how LOSS_CCE is considered in MQL5 (if it works at all). People have already asked in this thread for explanations about the implementation of all the loss functions specified in the ENUM_LOSS_FUNCTION enumeration, but there was no answer. And if, for example, LOSS_BCE is counted, then for LOSS_CCE I always get 0, no matter what matrix I put in. In short, it doesn't work with any of the tips - keras, numpy.

Here are the sources in MQL5. With the sources where the calculation formulas are taken from.

[Renat Akhtyamov]

Slava #:
Here are the sources in MQL5. With the sources where the calculation formulas are taken from.

vector,

probably the most useful thing in trading

[Stanislav Korotky]

Slava #:
Here are the sources in MQL5. With the sources where the calculation formulas are taken from

How does this relate to the matrices and examples of "parallel" platforms I gave above?

For example, I take matrices from the link to keras, call for them:

   matrix y_true = {{1., 0., 0.}, {0., 1., 0.}, {0., 0., 1.}};
   matrix y_pred = {{.9, .05, .05}, {.5, .89, .6}, {.05, .01, .94}};
   Print(y_true.Loss(y_pred, LOSS_CCE));
   Print(y_true.Loss(y_true, LOSS_CCE));

and get zeros.

[mytarmailS]

I'm completely out of the loop, but I'm wondering if a price can be represented by a graph, and if it has any advantage over the usual two-dimensional representation.

Who knows about this?

[Maxim Kuznetsov]

mytarmailS #:
I'm completely out of the loop, but I'm wondering if a price can be represented by a graph, and if it has any advantage over the usual two-dimensional representation.

Who knows about this?

If you have an imagination, you can represent the price with a bipartite graph. But it is better to imagine that she is a lady at heart :-)

[mytarmailS]

Maxim Kuznetsov #:

if you have imagination, imagine the price as a bipedal count. But it is better to imagine that she is a lady at heart :-)

An earl has no shares.

[Maxim Kuznetsov]

mytarmailS #:
The count has no shares.

That leaves "lady in the shower.")

[mytarmailS]

Maxim Kuznetsov #:

so that leaves "lady in the shower.")

Are you drunk?

[Maxim Dmitrievsky]

mytarmailS #:
I'm completely out of the loop, but I'm wondering if a price can be represented by a graph, and if it has any advantage over the usual two-dimensional representation.

Who knows about this?

Are Markov chains already boring? )

The failure of MO algorithms in trading is quite obvious and is not related to their type.

the prediction problem is incorrectly set.

[mytarmailS]

Maxim Dmitrievsky #:

1)Are the Markov chains getting boring yet? )

2) the failure of MO algorithms in trading is quite obvious and is not related to their type.

3)the prediction problem is incorrectly posed

1)this is different

2)The type of MO may not matter but the way information is presented matters.As an example CNN's boom compared to past AMOs.

3)those are my words)