This is my third project on Machine Learning. This project is made for educational purpose. This is the task from "Сириус - Курсы по машинному обучению". Task was to create a Machine-Learning LogisticClassification model
There is a csv file, that represents information (factors) that has some hidden relation to "satisfaction" of the customers.
Task required me to create a LinearLogisticClassification model in order to predict if customer is satisfied with his flight. All the factors were created and collected by a company, that made up a test for its customers
I have discovered the most useful numeric and categorial factors in order to optimize the model of classification. All parameteres are scaled.
Those are RESULTS (accuracy/precision/recall):
BOOTSTRAP BOUNDARIES (доверительные интервалы) of variable accuracy/precision/recall:
