How kushalj123/kickstarter-prediction works
This is a Google Colab notebook project for predicting Kickstarter campaign success from pre-launch features. Its documented workflow reads April and May 2021 CSV collections from Google Drive, cleans and merges each month, engineers features, trains an XGBoost classifier, evaluates it, and saves predictions. The README supports these stage relationships, but no notebook source excerpts were available to verify finer-grained calls or implementation details; all repository components therefore point to the notebook path.
Main entry points: ML_miniproject_PES1UG23CS319_PES1UG23CS343.ipynb Clean and merge
Components 9
Data preparation
- Clean and merge
ML_miniproject_PES1UG23CS319_PES1UG23CS343.ipynb - Cleaned monthly data
ML_miniproject_PES1UG23CS319_PES1UG23CS343.ipynb - Feature engineering
ML_miniproject_PES1UG23CS319_PES1UG23CS343.ipynb
Modeling and evaluation
- XGBoost training
ML_miniproject_PES1UG23CS319_PES1UG23CS343.ipynb - Model evaluation
ML_miniproject_PES1UG23CS319_PES1UG23CS343.ipynb
Saved outputs
- Prediction CSV
ML_miniproject_PES1UG23CS319_PES1UG23CS343.ipynb
Other components
- Notebook user
- Google Drive
- Monthly CSVs
Connections 9
- Notebook user → Clean and merge: runs notebook
- Google Drive → Monthly CSVs: provides CSVs
- Clean and merge → Monthly CSVs: reads and cleans
- Clean and merge → Cleaned monthly data: writes monthly data
- Cleaned monthly data → Feature engineering: provides data
- Feature engineering → XGBoost training: provides features
- XGBoost training → Model evaluation: evaluated by
- XGBoost training → Prediction CSV: saves predictions
- Prediction CSV → Google Drive: stored in