Inspiration
Customer churn prediction
What it does
classifies customer chrurn
How we built it
Using Machine Learning and Deep learning. With Integration of MLFlow
Challenges we ran into
For model performance Improvement(Data manipulation) normalized the features using log normal distribution but the performance didn't increase and then tried Generated Data using SMOTE and then trained model in the large data but still the accuracy remained same.
For model performance Improvement (Model training) Used complex Algorithms - GradientBoostingClassifier , XGBoostClassifier , CatBoostClassifier , AdaBoostClassifier , RandomForestClassifier to easy algorithm like Logistic Regession and Also trained Deep Neural Network with different weight Initializers , activation function ,input nodes and optimizer but models performance not Improved .
Accomplishments that we're proud of
The model is successfully working
What we learned
MLFlow and Dagshub
What's next for Customer Churn Prediction
read data from mondoDB deploy the model in AWS
Built With
- deep-learning
- machine
- machine-learning
- natural-language-processing
- python
Log in or sign up for Devpost to join the conversation.