Spaces:
Sleeping
Sleeping
Update README.md
Browse files
README.md
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
---
|
| 2 |
title: AppUserClustering
|
| 3 |
-
emoji:
|
| 4 |
-
colorFrom:
|
| 5 |
-
colorTo:
|
| 6 |
sdk: docker
|
| 7 |
app_port: 8501
|
| 8 |
tags:
|
|
@@ -12,9 +12,35 @@ short_description: User segmentation with KMeans clustering
|
|
| 12 |
license: mit
|
| 13 |
---
|
| 14 |
|
| 15 |
-
#
|
| 16 |
|
| 17 |
-
|
| 18 |
|
| 19 |
-
|
| 20 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
title: AppUserClustering
|
| 3 |
+
emoji: 👥✨
|
| 4 |
+
colorFrom: indigo
|
| 5 |
+
colorTo: gray
|
| 6 |
sdk: docker
|
| 7 |
app_port: 8501
|
| 8 |
tags:
|
|
|
|
| 12 |
license: mit
|
| 13 |
---
|
| 14 |
|
| 15 |
+
# 👥✨ KMeans User Segmentation
|
| 16 |
|
| 17 |
+
This Streamlit app performs **user segmentation** using a trained **KMeans clustering model**.
|
| 18 |
|
| 19 |
+
## What it does
|
| 20 |
+
- Loads a saved **KMeans model** (`kmeans_user_segmentation.pkl`) and **StandardScaler** (`scaler.pkl`)
|
| 21 |
+
- Accepts input via:
|
| 22 |
+
- **CSV upload**, or
|
| 23 |
+
- **Manual input (demo)**
|
| 24 |
+
- Assigns each row to a **cluster**
|
| 25 |
+
- Shows:
|
| 26 |
+
- cluster distribution
|
| 27 |
+
- cluster profile (mean feature values)
|
| 28 |
+
- optional PCA 2D visualization
|
| 29 |
+
- Allows downloading the results as CSV
|
| 30 |
+
|
| 31 |
+
## Required CSV columns
|
| 32 |
+
Your CSV must contain these columns exactly:
|
| 33 |
+
- `Average Screen Time`
|
| 34 |
+
- `Average Spent on App (INR)`
|
| 35 |
+
- `Ratings`
|
| 36 |
+
- `Last Visited Minutes`
|
| 37 |
+
- `New Password Request`
|
| 38 |
+
|
| 39 |
+
## Files in this Space
|
| 40 |
+
- `app.py`
|
| 41 |
+
- `kmeans_user_segmentation.pkl`
|
| 42 |
+
- `scaler.pkl`
|
| 43 |
+
- `requirements.txt`
|
| 44 |
+
|
| 45 |
+
## Note
|
| 46 |
+
Clustering is unsupervised: clusters represent **groups of similar behavior**, not ground-truth labels.
|