adelorenzo/stable
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

add shuffle parameter explaination

Browse Source
This commit is contained in:
longyu 2022-07-11 23:36:01 +02:00
parent bb1ab0be2a
commit 3fefa4b144
1 changed files with 7 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
docs/freqai.md
View File

@ -408,6 +408,11 @@ It is common to want constant retraining, in whichcase, user should set `live_re
### Controlling the model learning process
Depending on what AI model to be used, these parameter names could be different. For example, the accepted parameters for the `Catboost`
models are `data_split_parameters`, `n_estimators` and etc. For the model like SVM regression model, the accepted parameters are different.
Here we explan the parameters of `model_training_parameters` for `Catboost`:
The user can define model settings for the data split `data_split_parameters` and learning parameters
`model_training_parameters`. Users are encouraged to visit the Catboost documentation
for more information on how to select these values. `n_estimators` increases the
@ -425,6 +430,8 @@ where $W_i$ is the weight of data point $i$ in a total set of $n$ data points._
Finally, `period` defines the offset used for the `labels`. In the present example,
the user is asking for `labels` that are 24 candles in the future.
Note: since we work time series data and want to train a AI model to predict the future, the validation/test data should be the "future" by a given training data. Thus, we strongly recommend to disable `shuffle` parameter during the cross-validation steps. For more detailed explaination, visit [here](https://medium.com/@soumyachess1496/cross-validation-in-time-series-566ae4981ce4).
### Removing outliers with the Dissimilarity Index
The Dissimilarity Index (DI) aims to quantify the uncertainty associated with each