Skip to content

Latest commit

 

History

History
10 lines (6 loc) · 858 Bytes

File metadata and controls

10 lines (6 loc) · 858 Bytes

#Classification alghoritms comparison

The work aims to compare classification algorithms performance (Decision Trees, SVM, Logistic Regression, Bagging Classifier, Random Forest) without and with optimiziation. GridSearch and in some cases also graphical methods were used to find the best algohirtms parameters. Models were tested on Titanic datasets and the goal was to predict if passenger survived or not (binary classification) based on available information. The methodology of operation is described in detail in the notebook.

Using the best model, we can say with certainty level of ~ 83,3 % if Titanic passenger survived or not.

Alghoritms accuracies summary: image

This task can be developed by using other alghoritms like naive Bayes.