jamesor19/knowldge-graph-embedding-model-for-customer-churn-prediction

Project on the use of Knowledge Graph Embeddings for use for a classification problem for Capstone Project submitted for the MSc of Data Analytics

Knowldge Graph Embedding Model for Customer Churn Prediction

Capstone Project Master of Data Analytics NUIG

James O'Rourke 15334121

Requirements

• ampligraph version 1.4.0
• tensorflow version 1.15.0
• sckit-learn version 0.24.2
• numpy version 1.20.3
• imbalanced-learn version 0.8.0

The contents of the files are as such

• dataPrep : Juypter notebook with data preperation and creation of knowledge graph
• experiment : Juypter notebook contains the creation, traning and predcition of the KGE models
• Baseline Model : Juypter notebook contains the traning, predcition and evaulation of the Random Forest Model
• eval : Juypter notebook contains the evaulation of the KGE models

The data provided

• bank.csv and Telco_customer_churn.csv contains the orginal data
• bank_prepped.csv and Telco_prepped.csv contain the data after balancing the classes and preperation
• BankTriples.csv and TelcoTriples.csv contain the Knowldege Graph created from the above data