-
The source files consist of 2 parts.
-
The first part is 2022_02_20_churn_summative_part_1.ipynb
-
The input file required for part I is cell2celltrain_Small_6k.csv
-
The second part is 2022_02_26_churn_summative_part_2.ipynb
-
The input file required for part II are:
-
- df_imputed.csv
-
- features_selected_new.txt
-
The formal written report is report.pdf.
-
The requirement of this project is in Assessment Brief.pdf.
-
Supervised learning algorithm was used to build churn prediction model to help solve a telecoms company's customer churn problem.
-
Decision tree classifiers and optimisation techniques were used for feature selection.
-
The genetic algorithm was applied to a telecoms customer dataset consisting of 6380 rows and 57 features.
-
The Python programming language, Jupyter notebook and scikit-learn python package were used.