Neuro-forecasting of financial series (based on one article) - page 12

 
Break out the corners, please.
 
alexeymosc:
Break out the corners, please.

No problem. Anyway, from communicating with mersi no useful information, at least for me (I'm not responsible for others). And therefore, can easily blacklist him and do not respond to his posts.

My own offtops in the posts too, I think I'll remove, to make the topic more informative.

 

everyone here knows your gold dust and ai. they are an exact replica of your intelligence.

So shove all your knowledge and skills up your ass

and be humble.

 

This is a good example of what NS is capable of.

At first I optimized input data by maximal number of deals multiplied by profit, then this optimized indicator was fed to the input of the neural network. As a result, it improved the result can be said conditionally...... Yes, it got better, but not as much as it should....

 
alexeymosc:

In my opinion, a very high quality article, the main pitfalls are outlined, the conclusions are not startling but only slightly surprising. And the ideas are quite progressive (meaningful, ideologically fuelled, nurtured...)

It's all bullshit. Theory breaks down in practice. The principle should be transparent. Like comparisons, concrete results from the past. Everything else, odd logic... that's what fuzzy logic is. You go to the shop and do not pay there with your fuzzy logic, in the shop you will be asked exactly in grams, and in local currency
 
eugene-last:
It's all bullshit. Theory breaks down in practice. The principle should be transparent.

These are academic articles. These people do not have to present you with the history of the trade.

The articles are for the purpose of enriching the community with ideas, nothing more.

Reason: