parkingSegmentation

parking users segmentation from the parking system operational log

1. Preprocess.py Adds new variables to original log dataset Input file: history_origin.csv Output file: history_2022_02_preprocessed_.csv

2. Exploratory_history_22_02.py Cleans the data and creates Users dataset Input file: history_2022_02_preprocessed_.csv Output files: users_2022_02.csv + history_2022_02_to_powerbi.csv

3. Exploratory_users_22_02.py Cleans the data and does some exploratory analysis Overwrites the file users_2022_02.csv Input file: users_2022_02.csv

4. scale_and_pca.py Scales data and applies Principal Component Analysis Input file: users_2022_02.csv Output file: explained_var.csv

5. kmeans.py Clustering with KMeans from sklearn.cluster package Input file: users_feb_pca.csv Output files: users_2022_2_labeled.csv + centers_kmeans6.csv

6. hierarchical.py Clustering with AgglomerativeClustering from sklearn.cluster package Input file: users_2022_2_labeled.csv Output file: users_2022_2_labeled.csv

7. som.py Clustering with minisom from MiniSom package Input file: users_2022_2_labeled.csv Output files: users_2022_2_labeled.csv + centers_som7.csv

8. fraud.py Anomaly detection using SOM Maps Input file: users_2022_2_labeled.csv Output file: users_fraud_20.csv

Also uploaded DBSCAN and Gaussian Mixture models code, although they don't fit the data