EDA & Machine Learning exercises

❤️‍🔥 OKcupid
- Dataset : OKcupid Profiles
- Mini EDA exercise : OKcupid EDA
- User segmentation : OKcupid Customer segmentation
  - Lessons learned : The concept of clustering is more or less clear enough to apply. However, I am still having difficulties in understanding how to interpret the result of clustering algorithms 🤷🏻‍♀️ Higher number of featrues seems to make more difficult to cluster datapoints and visualize the result.
🚴🏻‍♂️ Stolen bicycle report in Berlin Germany
- Dataset : https://www.kaggle.com/datasets/staefff/stolen-bike-report-berlin
- EDA exercise: Berlin stolen bicycle report
- Lessons learned: shame that I couldnt do forecasting because there were many missing dates within the period. One year dataset will lead to overfitting with high possibility.
- Tableau visualization:
- Shared in Kaggle: Kaggle
🐺 Woes Club Churn Prediction with Pycaret
- Dataset: Woes Club
- References: Pycaret
- Churn Prediction Model
- Lesson learned: Pycaret needs certain environment settings. It is definitely a useful package to easily get the best model for dataset. But I would not totally rely on the package. Always try out different approaches for data preprocessing and model building.
🚢 E-Commerce Lieferverzug-Prognose
- Dataset: Product Shipment Delivered on time or not? To Meet E-Commerce Customer Demand
- Lieferverzug-Prognose

Name		Name	Last commit message	Last commit date
Latest commit History 38 Commits
Lieferverzug_Prognose_.ipynb		Lieferverzug_Prognose_.ipynb
OKCupid_Profiles_EDA-2.ipynb		OKCupid_Profiles_EDA-2.ipynb
OKCupid_UserSegmentation_ClusteringAlgorithms2.ipynb		OKCupid_UserSegmentation_ClusteringAlgorithms2.ipynb
README.md		README.md
Stolen_Bike_Report_in_Berlin.ipynb		Stolen_Bike_Report_in_Berlin.ipynb
e92665fb.xlsx - Table 1.csv		e92665fb.xlsx - Table 1.csv
stolen_bike_report_in_berlin.ipynb		stolen_bike_report_in_berlin.ipynb
woes_club_churn_prediction_withPycaret.ipynb		woes_club_churn_prediction_withPycaret.ipynb

Provide feedback