Baseline Model trained on Airlinesuiztcxpg to apply classification on Delay
Metrics of the best model:
accuracy 0.612210
average_precision 0.405509
roc_auc 0.635865
recall_macro 0.594188
f1_macro 0.569725
Name: LogisticRegression(C=0.1, class_weight='balanced', max_iter=1000), dtype: float64
See model plot below:
Pipeline(steps=[('easypreprocessor',EasyPreprocessor(types= continuous dirty_float low_card_int ... date free_string useless
Airline False False False ... False False False
Flight True False False ... False False False
AirportFrom False False False ... False True False
AirportTo False False False ... False True False
Time True False False ... False False False
Length True False False ... False False False[6 rows x 7 columns])),('logisticregression',LogisticRegression(C=0.1, class_weight='balanced',max_iter=1000))])
In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.
Pipeline(steps=[('easypreprocessor',EasyPreprocessor(types= continuous dirty_float low_card_int ... date free_string useless
Airline False False False ... False False False
Flight True False False ... False False False
AirportFrom False False False ... False True False
AirportTo False False False ... False True False
Time True False False ... False False False
Length True False False ... False False False[6 rows x 7 columns])),('logisticregression',LogisticRegression(C=0.1, class_weight='balanced',max_iter=1000))])EasyPreprocessor(types= continuous dirty_float low_card_int ... date free_string useless Airline False False False ... False False False Flight True False False ... False False False AirportFrom False False False ... False True False AirportTo False False False ... False True False Time True False False ... False False False Length True False False ... False False False[6 rows x 7 columns])
LogisticRegression(C=0.1, class_weight='balanced', max_iter=1000)
Disclaimer: This model is trained with dabl library as a baseline, for better results, use AutoTrain.
Logs of training including the models tried in the process can be found in logs.txt