Machine learning in trading: theory, models, practice and algo-trading - page 1738
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
Here, what to do with it? I'm not very good at clustering yet
predict on new cluster data
In R this is done through the predict function in Python, I don't know.
predict on new cluster data
In R, this is done through the predict function in Python, I don't know.
how to get the cluster number with this matrix and the fetch values
how to get the cluster number, having this matrix and feature values
Or if you want to do it in another program
then save the matrix with centroids into the new program and when new data come into the program then run the data through the matrix (line by line) for proximity (Euclidean, most likely), which centroid will be closest to the new data, this will be the cluster number
Or if you want to use another program
then save the matrix with centroids into a new program and when new data come into the program then run the data through the matrix (line by line) for proximity (Euclidean, most likely), which centroid will be closest to the new data, it will be the cluster number
Thanks, I'll read it.
in other algorithms cluster. the same scheme, or more complicated?
Thanks, I'll read it.
In other algorithms, is the clust. following the same scheme, or more complicated?
the same, the vast majority
listen, question ))))
If you didn't know how to recognize new data by clustering, where did you get the "cluster test" for the test?
Are these the same clusters that were involved in the training?
listen, question ))))
if you didn't know how to recognize new data by clustering, then where did you get the "cluster test" for the test?
Are these the same clusters that were involved in the training?
there's a prefix there... i'm just too lazy to go into the code and decipher how it counts
there's a prefix there... I'm just too lazy to go into the code and decipher how it counts
those clusters from the test were not involved in the training?
because if they did, you know))those clusters from the test were not involved in the training?
because if they did, you know))No, of course not.
Now the whole irony of a trading strategy that relies on frequency transforms, which is when we get a perfect circle at the end of the futures, there and any fool can build it. You just get the last week of the last well-functioning TS, and at the beginning you have to almost in the morning, and then even in the afternoon. It's work, well, if I had the opportunity at least once to run and look for these figures. I would be glad.
When you get lower-quality figures, different from the first one, then you should do the following, I leave the authorship to myself.
It is necessary to ignore the first signal based on the concept that you need to build two strategies diametrically, when you have two arrows simultaneously up and down on the first signal. Schröding's cat, the concept of quantum or stupid state "I do not know", well, fuck it, but you find out exactly what is still the cycle now, stupidly its direction and therefore the first signal will be directing, and you know why I took it????
Because on the indented signal, I'm a sly one, not like some people, if I tried, I could not get a better circle. That's the market right now. And the circle without the indentation is just really cooler than the circle of the previous signal. I'm just waiting for the second and third signals, which in theory will work. The first one, I know exactly the direction of the found cycle, but I do not know if the cycle I have found is correct in principle. And in the process of building it, it is constantly changing. However, we're not talking specifically about cycles, we're talking about the basic application of cycles in the context of CSSA.
From the story: Back in the olden days, the traders got bored and came to the radio amateurs in the laboratory and asked, "What do you know about frequency oscillations? OOO, answered the shaggy-haired one with the moustache and the wild engineer's look, I know everything. Look at the oscilloscope lying in the corner :-) that's how digital filters in stock trading came about. The famous FATL-SATL. I understand SSA is an offspring of these filters with CSSA as an apogee. No better at the moment. Slightly worse than a crawler.
So, the quality of the obtained circle directly depends on the time of futures expiration and the current volatility. That is, having received a circle, it is necessary to determine its quality relative to the specified variables. The criterion of quality can only be determined empirically. That's what I'm talking about here. If you have official data on the studies conducted, I ask you to inform me. I do not need the offended looks of losers.
This is if you start this study and finally put the point on the "I".
Because if you get qualitative circles for the current parameters of aspiration and volatility, you steadily earn, and maybe not.
That's just a way of putting it. I'm so excited %-)