ACTIONS news, forecasts, expectations 2022 - page 53

[Yuriy Zaytsev]

JRandomTrader #:

Not exactly emotion. I have learned the most difficult thing - not to trade.

I think that robots that work by themselves don't exist, but I think that a robot that was told to sell today or this week or this month, but sell and fix it, so it fulfills all the recommendations.

I.e., a semi-automatic robot is a real working thing.

..

The robot can be sharpened on the last trend, say, the semi-annual one... I.e. it is good to optimize it in the tester on the last history.

I managed to do it in 2007, when I participated in the Championship and even got 6th place in the world.

But in the beginning of 2008 my trading robot has already started to fall apart.

[Yuriy Zaytsev]

By the way, according to the rules, the robot had to be submitted to MQ 10 days before the Championship, i.e. by September 20.

I sent the debugged and customized version a few hours before the deadline, I was fine-tuning and optimizing the next version, but I didn't have time,

but I didn't have time to upload it, literally a day was not enough.

When I ran it on real data after the championship, I've found out that it has reached over 60k and would be the second in 2007 :)

My first robot was a betting trader with cosmic profitability and it raised 160k for 3 months from 10k. And the second place was taken by the American expert wakena who got 50k.

Mine was on divergence + very good fitting + lucky with the trend, I got only 6 in the world.

I sharpened the bias to buy, at sharp sales the robot kept trades longer, at purchases shorter.

[Fast235]

Yuri, who won there again?

[Fast235]

ideally comments on the first three.

[JRandomTrader]

Yuriy Zaytsev #:

In my opinion, robots that work by themselves do not exist, but the robot that was told to sell today or this week or this month, we just sell and fix it, it fulfills the recommendation correctly, it is real.

I.e., a semi-automatic robot is a real working thing.

..

The robot can be sharpened on the last trend, say, the semi-annual one... I.e. it is good to optimize it in the tester on the last history.

I managed to do it in 2007, when I participated in the Championship and even got 6th place in the world.

But, this robot already fell apart at the beginning of 2008.

I've already shown you the result of a robot that has been trading for three years without any changes. There is no super yield, but it can work.

https://www.mql5.com/ru/forum/387540/page5#comment_27381923

[Yuriy Zaytsev]

Fast235 #:
Yuri, who won there again?

in 2007, Batter, a guy from the Ukraine.

neural network from 10k to 160

There was a big interest in neural networks back then.

I communicated with him once by mail and in person at MQL.

He had been dealing with neural networks for 10 years before that. By the way, his neural network was leaked in 2008 :) too.

I felt sorry for putting a decent period of my life into studying neural networks :)

[Yuriy Zaytsev]

JRandomTrader #:

I have already shown the result of a robot that has been trading for 3 years without change. No super returns, but it can do something.

https://www.mql5.com/ru/forum/387540/page5#comment_27381923

And the trend is not set ?

[JRandomTrader]

Yuriy Zaytsev #:

and the trend is not set?

No. And there is also a complete symmetry of conditions for long and short. It is only since 3.22 that I started to move away from symmetry and combine it.

Moreover, the same robot with exactly the same settings trades not only RTS, but also MIX (more or less successfully) and Si (not so successfully).

[Yuriy Zaytsev]

JRandomTrader #:

No. And in this one there is a complete symmetry of conditions for long and short. It is only since 3.22 that I started to move away from symmetry and combine it.

Moreover, the same robot with exactly the same settings trades not only RTS, but also MIX (more or less successfully) and Si (not so successfully).

It is really interesting!

I even started another threadhttps://www.mql5.com/ru/forum/388746

[Fast235]

JRandomTrader #:

No. And in this one there is a complete symmetry of conditions for long and short. It is only since 3.22 that I started to move away from symmetry and combine it.

Moreover, this same robot with exactly the same settings trades not only RTS, but also MIX (more or less successfully) and Si (not so successfully).

Reminds me of our favourite, Maxim's, how much was thrown at him)

networks, clearly not in these years.