Discussion of article "Evaluation and selection of variables for machine learning models" - page 3

[Vladimir]

Vladimir Perervenko:

First, the definition of Classification is given at the kindergarten level. Then it is told that uncertainty is generated(!?) And ends as always: "Where is the key to the flat where the money is?".

You need more theoretical training. Study, study and study again ... You know.

And be more modest.

PS. Put your proposal in Freelance. Get a real product.

It's not the first time I've heard about the flat key. Re-read my proposal:"The details of the model, data normalisation, and their selection do not interest me. I am interested in the results of predictions ..." And predictions on the past history from 2000 to today. So I guess we all can't read. In short, the theory continues. We re-read other books and articles and wrote our own. At least try to trade on the real market using your own methods and then write articles. All right, academic theorists. I gave you some advertising here, and the branch started to drown in the flood of new articles.

[Sergey]

Here are a couple more articles on the subject:

http://robotwealth.com/machine-learning-financial-prediction-david-aronson/

http://robotwealth.com/machine-learning-for-financial-prediction-experimentation-with-david-aronsons-latest-work-part-2/

Perhaps someone will find it useful.

[madpower2000]

Andrew Ng - in Russian transcription it's probably more correct Andrew Eun ;)

Firstly, a HUGE thank you to the author for a unique series of articles, this is really the Grail for those who are in the subject or at least trying to understand!

Secondly, without prior training reading these materials for the average person is a dark forest, I would even say lies beyond his event horizon of cognitive abilities. Therefore, it is highly recommended to listen to courses on machine learning at econometrics and linear algebra from the Higher School of Economics on the same resource.

And only after that perhaps the understanding of what Vladimir writes about in his articles will come.

Once again, BRAVO!

[mg64ve]

Dear Vladimir, thanks for your nice article.

I am reading all your articles and they are very interesting.

Regarding this one, I still do not understand very well why do you use preProcess to partition data and how are data splitted?

From my experiments this function splits data into a different order.

The question is: how can I restablish the original order after I have the results from the predict function?

It appears this result it is in a different order.

Thanks in advance for your comments.

Cheers 

[Vladimir Perervenko]

mg64ve:

Dear Vladimir, thanks for your nice article.

I am reading all your articles and they are very interesting.

Regarding this one, I still do not understand very well why do you use preProcess to partition data and how are data splitted?

From my experiments this function splits data into a different order.

The question is: how can I restablish the original order after I have the results from the predict function?

It appears this result it is in a different order.

Thanks in advance for your comments.

Cheers 

Hi,

You probably misunderstood. Split  of the functions holdout() performed. Then the function preProcess() defined by the parameters of normalization in the training set. And then the code ..

Best regards

Vladimir

> idx <- rminer::holdout(y = data.f$Class)
> prep <- caret::preProcess(x = data.f[idx$tr, -ncol(data.f)],
+             method = c("spatialSign"))
> x.train <- predict(prep, data.f[idx$tr, -ncol(data.f)])
> x.test <- predict(prep, data.f[idx$ts, -ncol(data.f)])
> y.train <- data.f[idx$tr, ncol(data.f)]
> y.test <- data.f[idx$ts, ncol(data.f)]

[JunCheng Li]

Is that worng?  " best <- Cs(cci, cmo,  slowD, oscK, signal, tr, ADX. chv, atr, ar)"

 

[JunCheng Li]

Should the right code is:

>library(Hmisc)

>best <- Cs(cci, cmo,  slowD, oscK, signal, tr, ADX,chv, atr, ar)

the "ADX." should be "ADX,"?

[Vladimir Perervenko]

JunCheng Li:

Should the right code is:

>library(Hmisc)

>best <- Cs(cci, cmo,  slowD, oscK, signal, tr, ADX,chv, atr, ar)

the "ADX." should be "ADX,"?

Да, это опечатка в статье.

[Zhi Long Yang]

Thank you very much to the author of the article. I've just started and I'm having a problem. I installed RStudio, not Revolution R Open 3.2.1 as suggested by the author. The "RandomUniformForests" package andthe "RoughSets" package are loaded, but the nearZeroVar() function and the findLinearCombos() function are not called correctly. The "RandomUniformForests" package and the "RoughSets" package are loaded, butthe nearZeroVar() function andthe findLinearCombos() function are not called correctly, are these functions specific to Revolution R Open?

[JulInParis]

Hi Vlad,

I'm trying to rerun your example step by step.

In the section Input data , The In(p=16) function deals with a price object. What is its R- format or class ( zoo, xts or dataframe ) and how does it look like ( its column names, etc..). Without these information, it's impossible to run the command    x <- In(p = 16) ...

 

Best regards.

 

Julien