Sleep Health and Lifestyle Clustering using Scikit-Learn

---

Data

https://www.kaggle.com/datasets/informateur234/sleep-health-and-lifestyle-dataset

---

Data Distribution

After preprocessing, the data distribution are shown in the images above. Applied linear discriminant analysis(LAD) for each class to be seperated which will be helpful for the algorithm to learn. Oversampling for train set is used.

---

Clustering Models

Selected Models are the following;

1. KMeans Clustering
2. MeanShift Clustering
3. Gaussian Mixture Clustering
4. DBSCAN Clustering

---

Clustering Result Before Hyper-Tuning

Clustering Result After Hyper-Tuning

---

Model Analysis

KMeans:

MeanShift:

Gaussian Mixture:

DBSCAN: