0
$\begingroup$

I have an OHLCV* dataset that starts on 01-01-2000 and ends on 31-12-2003 and I want to evaluate a model, say an SVM regressor. In other words, given some daily features describing the dynamics of the price for that day, I want to predict the price for the next day.

What is the correct routine to evaluate the performance of the model from 01-01-2003?

These are the steps I performed:

For testday in [01-01-2003, 31-12-2003]:
1. split the dataset into a training set from 01-01-2000 to testday-1 and a test set of the single testday vector
2. do standardization/normalization on the training set and test set
3. train the model on the training set
4. test the model on the testday (i.e., the day after the last day of the training set)
5. add the prediction an a vector of daily predictions

Evaluate the accuracy on the predictions made

Is this routine of training on a increasing number of days and testing on the next day correct?

*Each sample of the dataset is a daily vector of the Open price, Highest price, Lowest price, Close price, and Volume for the asset under analysis.

Improve this question
$\endgroup$
2
  • 2
    $\begingroup$ OHLCV says nothing for many a statistician / machine learner. Consider spelling the acronym out explicitly. Thank you. $\endgroup$
    – Richard Hardy
    Commented Sep 16 at 9:33
  • 1
    $\begingroup$ Hi. In addition to Richard Hardy's comment, consider that there are people who may be able to answer your question, but who do not know or use Python. Also consider other people who don't know Python and could have the same question as you but in a different programming language, and who would be interested in an answer. You should rewrite your question to make it language-agnostic, i.e. describing in plain English what your code is doing. Imagine if someone had the same question as you in a language like C or Haskell, and did not provide any explanation as to what their code is doing. $\endgroup$
    – J-J-J
    Commented Sep 16 at 10:41

0

Reset to default

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.