Discussion of article "Machine learning in Grid and Martingale trading systems. Would you bet on it?" - page 4
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
That was the goal, to find other setups that are impossible without the mesh
Yes, Maxim, I realised that when you were about to write about it ).
Speaking of crises, we need a separate definition of crisis states. Put fundamental and something else on the inputs. It seems that even one VIX has a high predictive power.
In general, I think that ensembles are a powerful thing, not in the classical sense, when an ensemble of weak models solves one problem strongly, but in the fact that the members of the ensemble solve different problems, the answers to which help to make decisions of the final TS. Thus, one can teach to identify unstable states in the market, when patterns break, by different narrow sources - oil market, housing market, debt market, etc.
Reducing risk or shutting off trading altogether during these periods should increase resilience. I also assume after this period many patterns require rethinking/re-learning.
Yes, Maxim, I realised that when you were about to write about it )
Speaking of crises, we need a separate definition of crisis states. Put fundamental and something else on the inputs. It seems that even one VIX has a high predictive power.
In general, I think that ensembles are a powerful thing, not in the classical sense, when an ensemble of weak models solves one problem strongly, but in the fact that the members of the ensemble solve different problems, the answers to which help to make decisions of the final TS. Thus, it is possible to teach to determine unstable states in the market, when patterns break, by different narrow sources - oil market, housing market, debt market, etc.
Reducing risk or shutting off trading altogether during these periods should increase resilience. I also assume after this period many patterns require rethinking/re-learning.
For example, the 2008 crisis and its echo in 2015 was a surprise, the fundamentals there were not particularly revealing. Rather, macroeconomic analyses of the situation. So it is very difficult to build such a macroeconomic model with IO. It has to analyse sentiment, tweets, google queries and so on then to pull out the context of the overall market sentiment. I don't have a background in NLP.
And it's easier to play with models, post boosts and such. I want to train some models for different periods and compose.
For example, the crisis of 2008 and its echo in 2015 were a surprise, the fundamental was not particularly revealing there. Rather, macroeconomic analysis of the situation. So it is very difficult to build such a macroeconomic model with the help of IO. It has to analyse sentiment, tweets, google queries and so on then to pull out the context of the overall market sentiment. I don't have a background in NLP.
And it's easier to play with models, post boosts and so on. I want to train some models for different periods and compose.
Yes, I agree. and in 2008 and 2020 would not help. in 2008 rating agencies were drawing beautiful numbers. in 2020 it is a surprise for all forecasts (except for Gates perhaps :) ). Black swans can't be taken into account in the model.
But we can probably indirectly train the model to determine the beginning of a storm on the market, when it is better to stand aside.
But the question is whether for a particular strategy the moment when the storm starts coincides with the moment when the regularity exploited by the strategy stops working.
Yes, I agree. it would not have helped in 2008 and 2020 either. in 2008 rating agencies were drawing beautiful figures. in 2020 it will be a surprise for all forecasts (except for Gates perhaps :) ). Black swans can't be taken into account in the model.
But it is possible to indirectly train the model to determine the beginning of a storm on the market, when it is better to stand aside.
But the question is whether for a particular strategy the moment when the storm starts coincides with the moment when the pattern exploited by the strategy stops working.
Is it possible to make manual markup in the script? I would like to specify where to buy and where to sell.
Is it possible to make manual markup in the script? I would like to specify where to buy and where to sell.
I'll have to think about it, I can't tell you right away. We will have to take into account those data that are not marked manually. Most likely, it will be necessary to redo everything.
Or do it through a filter, as in the previous articleYou will have to take into account data that is not manually labelled in any way
This is not necessary, only the ones with my markup should be taken into account.
You must be a fisherman to write such a title. I haven't fished for a long time. It's a pity to eat. You could eat something else. It's a game, catching a fish. Then eat it. By the way, the Chinese invented soya fish a long time ago. It even has bones, but it's richer in calcium. But if you spice it up a bit, it's quite a good meal. My wife (by the way, she had a friend who was a chef of a large restaurant), when she cooked it for the first time, I could not believe that it could be cooked. Imagine, quite simple products, and I'm not a bad cook myself. But when I tried it at her place, at first I gawked, then I asked her for the recipe for a long time. I never believed it was possible to cook like that. But I've been eating it for years. I never got used to it, by the way. I always remember it and I think it was some kind of obsession. Magic. I think she was doing something completely wrong, or not what she was telling me.
The interesting thing about cooking, she said you could cook any piece of meat in different flavours. It's always different. And you can't tell the difference. And what's more, it's not the main ingredients, it's the spices. Of course, you can buy super expensive and exquisite products, spending a lot of time and money on this activity. But you can do the simplest thing. Just add spices. It may sound weird. But I still don't believe it works.
It's amazing how expensive foods can be replaced with cheap foods and simple spices. Strangely enough, despite what many would say is more no than yes. It works. What's more, it works even better. Why waste a lot of energy and time running around shopping for bluebird when you always have everything at your fingertips.
As for the technique, it's kind of good on one hand, but it looks kind of angular. Of course, you can show brutal, pumping in algorithms and performance. And maybe I'm wrong here, but everything is good in moderation. You should always look for a comprehensive picture. I think everything should be in moderation.
Of course, it is good to filter out the indicators with power, and why not, of course there will be results, but if you use more flexible approaches, then there may be unnecessary nets, and there is no reason to torture the fish. The stick is always about two ends. If on one side on a bunch of candles teraflops, then maybe on the other side you need at least a mashka.
Otherwise the picture will be very lopsided if we overlook something. That's my opinion. You can ignore it. But I think you agree that when we get caught up in one thing, we can overlook the other.
It's a good article, but it's a little abstruse. But that's okay. All in all, it's okay.
You must be a fisherman to write such a title. I haven't fished for a long time. It's a pity to eat. You could eat something else. It's a game, catching a fish. Then eat it. By the way, the Chinese invented soya fish a long time ago. It even has bones, but it's richer in calcium. But if you spice it up a bit, it's quite a good meal. My wife (by the way, she had a friend who was a chef of a large restaurant), when she cooked it for the first time, I could not believe that it could be cooked. Imagine, quite simple products, and I'm not a bad cook myself. But when I tried it at her place, at first I gawked, then I asked her for the recipe for a long time. I never believed it was possible to cook like that. But I've been eating it for years. I never got used to it, by the way. I always remember it and I think it was some kind of obsession. Magic. I think she was doing something completely wrong, or not what she was telling me.
The interesting thing about cooking, she said you could cook any piece of meat in different flavours. It's always different. And you can't tell the difference. And what's more, it's not the main ingredients, it's the spices. Of course, you can buy super expensive and exquisite products, spending a lot of time and money on this activity. But you can do the simplest thing. Just add spices. It may sound weird. But I still don't believe it works.
It's amazing how expensive foods can be replaced with cheap foods and simple seasonings. Strangely enough, despite what many would say is more no than yes. It works. What's more, it works even better. Why spend a lot of effort and time running around shopping for the bluebird when you can always have everything on hand.
As for the technique, it's kind of good on one hand, but it looks kind of angular. Of course, you can show brutal, pumping in algorithms and performance. And maybe I'm wrong here, but everything is good in moderation. You should always look for a comprehensive picture. I think everything should be in moderation.
Of course, it is good to filter out the indicators with power, and why not, of course there will be results, but if you use more flexible approaches, then there may be unnecessary nets, and there is no reason to torture the fish. The stick is always about two ends. If on one side of a bunch of candles teraflops, then maybe on the other side you need at least a mashka.
Otherwise the picture will be very lopsided if we overlook something. That's my opinion. You can ignore it. But I think you agree that when we get into one thing, we can simply overlook the other.
It's a good article, but it's a little abstruse. But that's okay. All in all, it's okay.
Many people know that nets can catch more fish, but they are prohibited. If you block the river with a net, it will block the course of a school of fish, and many fishermen will be left without food that evening. Of course, nowadays few people get food this way, mostly amateurs who do not mind catching a couple or three bream and fry them on a frying pan. Sometimes you sit in the boat for an hour or more, and still no bite. In addition, you are constantly snagging blades on snags. At such moments you regret that you did not put a net and did not deal with more pressing problems. Of course, fishing with a net is not the same as fishing with a trolley. Many, including aesthetic, moments are missed. A real fisherman, apparently, will never go for it. He wants to feel every movement, feel the wind in his face and listen to the buzzing of mosquitoes in coastal areas, interspersed with the croaking of frogs in dams. Such fishing gradually goes into the deep evening, when the sun at sunset illuminates a mirror-like motionless water surface, the taut string of which is disturbed here and there by rare bursts of feeding pike and pikeperch. This is definitely a special philosophy, inaccessible to the urbanised consumer who is used to taking ready-made products from the oven.
Many people know that you can catch more fish with nets, but they are forbidden. If you block the river with a net, it will block the course of a school of fish, and many fishermen will be left without food that evening. Of course, nowadays few people get food this way, mostly amateurs who do not mind catching a couple or three bream and fry them on a frying pan. Sometimes you sit in the boat for an hour or more, and still no bite. In addition, you are constantly snagging blades on snags. At such moments you regret that you did not put a net and did not deal with more pressing problems. Of course, fishing with a net is not the same as fishing with a trolley. Many, including aesthetic, moments are missed. A real fisherman, apparently, will never go for it. He wants to feel every movement, feel the wind in his face and listen to the buzzing of mosquitoes in coastal areas, interspersed with the croaking of frogs in dams. Such fishing gradually goes into the deep evening, when the sun at sunset illuminates a mirror-like motionless water surface, the taut string of which is disturbed here and there by rare bursts of feeding pike and pikeperch. This is definitely a special philosophy, inaccessible to the urbanised consumer who is used to taking ready-made products from the oven.
That's the way it is. If you have a herd, it is better to milk a couple of dozen cows to collect the cream, and sell the excess and buy new hay. A lot of people want everything at once. Your article is a guarantee for the majority. If you have a herd you need to move it, otherwise it will scatter. Infusion of a new portion of some guarantees is always praiseworthy. It is even difficult to object here. The most profitable thing is, will be, and will remain a good show. And better than a good show, there can only be a better show to meet the needs of the majority. That's right. The crowd rules. So you have to throw more and more fodder at the audience. The important thing is that everybody's willing to pay for what they get. Ready to have a steady income? Sign here. Then it's on to the seats you buy on the ship. The audience gets their seat in the audience.
Score, you don't even need rational logic. Come and get it. The more we buy, the more we take. The article is right on the money. You want to work the net, you work the net. By the way, I've got a couple of orders for mesh right now. Okay, the crowd rules. :)