Norbert Kuder
Norbert Kuder
Add models from smooth package in R.
### Describe the workflow you want to enable I suggest adding new transformers to scikit-learn named `LogTransformer` and `LogWithShiftTransformer`, which would add the functionality of applying a logarithmic transformation and...
### Describe the workflow you want to enable Consider to add ReliefF and RReliefF feature selectors to sklearn.feature_selection. ### Describe your proposed solution Add ReliefF and RReliefF feature selectors. ###...
Add built-in schema validation capabilities to pandera. It should check time series gaps based on datestamp gaps. Time series data typically requires continuous timestamps without gaps. Currently, users have to...
Currently, statsmodels only provides an interface to a very limited subset of X13-ARIMA-SEATS model features. There is a seasonal package in R that offers full support for this model, and...
Add the Hurst exponent calculation feature, providing a tool for analyzing long-term memory effects in time series data based on: [Hurst exponent](https://en.m.wikipedia.org/wiki/Hurst_exponent)
Currently, the STLForecaster is a rather limited solution that often fails to decompose time series effectively (residuals can still have the seasonalities). One possibility for improvement is to incorporate the...
Add the cross- and termporal forecasting based on R packages: [thief](https://github.com/robjhyndman/thief) [MAPA](https://github.com/trnnick/mapa) [FoReco](https://github.com/daniGiro/FoReCo) I am eager to help to implement those algorithms.
Since the statsmodels package does not provide the full access to the X-13ARIMA-SEATS model, we could add this as a feature to sktime. I am eager to help with this,...
Add the mRMR, ReliefF and RReliefF feature selectors like in Matlab: [mRMR](https://www.mathworks.com/help/stats/fsrmrmr.html) and [ReliefF and RReliefF](https://www.mathworks.com/help/stats/relieff.html)