Blog
dask, survival analysis
From Data Portals to Portals of Doom: Avoiding it with Dask
Circumvent the limitations of second-hand data portals: Orange with Dask enables you to process big data sets while ensuring the source's authenticity and enabling custom workflows.
Jaka Kokošar, Nov 08, 2023
dask, spectroscopy
Can Orange explore a 13 GB data set?
An experimental version of Orange supports much larger data sets. It never loads the whole data set into the working memory; data is only processed in small chunks. We show a case study on a 13 GB spectroscopy data set.
Marko Toplak, Oct 28, 2023
dask, development
Dask all Folks: Preparing Large Datasets
Preparing large HDF5 datasets that load into Orange as on-disk data.
Noah Novšak, Oct 24, 2023
conference, research
Recap of 26th International Conference on Discovery Science
Highlighting our papers from the International Conference on Discovery Science
Martin Špendl, Oct 20, 2023
update, video series
Fall Season Brings Fresh Content to the Introduction to Data Science Series
Updates in the Introduction to Data Science Video Series, new video logistic regression nomogram.
Erika Funa, Oct 17, 2023
fairness
Why Removing Features Isn't Enough
Find out why merely removing protected attributes will not fix bias. Features often correlate, letting models infer biases. Fairness algorithms are key for genuine bias mitigation.
Žan Mervič, Sep 19, 2023
fairness, reweighing
Orange Fairness - Reweighing as a preprocessor
Expanding on the Orange fairness Reweighing widget: using it as a preprocessor and integrating new fairness scoring metrics.
Žan Mervič, Sep 19, 2023
fairness, reweighing
Orange Fairness - Reweighing a Dataset
Introducing the Reweighing widget in Orange, a solution for dataset bias mitigation.
Žan Mervič, Sep 19, 2023
fairness, equal odds postprocessing
Orange Fairness - Equal Odds Postprocessing
Explore the Equal Odds Postprocessing widget in Orange, designed to fine-tune your model's fairness. We explain how the algorithm operates and showcase its effectiveness with an example using the German credit dataset.
Žan Mervič, Sep 19, 2023
fairness, adversarial debiasing
Orange Fairness - Adversarial Debiasing
Learn how to use the Adversarial Debiasing model in Orange for fairer machine learning.
Žan Mervič, Sep 19, 2023
fairness, dataset bias
Orange Fairness - Dataset Bias
Orange now supports methods for detecting and mitigating bias in machine learning.
Žan Mervič, Sep 18, 2023
dask, development
A dash of Dask
Updates in the dask project, integrating machine learning methods and enhancing efficiency in a data preprocessor.
Noah Novšak, Aug 31, 2023
education
Pumice challenge
One thousand Slovenian students took part in a data mining challenge
Anja Mejač, Janez Demšar, May 26, 2023