Script developed for looking for the best classifier for wine quality predictor. Based on kaggle implementation: https://www.kaggle.com/vishalyo990/prediction-of-quality-of-wine
Hi this is the first uploaded version of wine quliaty calssifier, based on dataset: https://archive.ics.uci.edu/ml/machine-learning-databases/wine-quality/
For running the code please download the repository, install necesarry modules and run the "Winko.py" script.
In the future, I plan to prepare software to classify it by quality scale and not just binary as It is now.
For nor (may 2020) F1 score for best algirthm (Random Tree) for predicting "bad wine" is more than 90% and for "good wine" around 65%.
New idea which is being developed is an idea to first detect outliers, such as best and worst wines and then classify them as good and bad. Test showed that this idea is not worth trying. Not any of the results present satisfactory result.
Feel free to use it.
for questions please write to: adam.k.mika@gmail.com