docs: add more information about the dataset

README.md

@@ -20,4 +20,40 @@ configs:
   - split: train
     path: 
     - "raw_exercise.csv"
----
+---
+
+# Calorie Burnt 15k
+
+A comprehensive dataset combining exercise and calorie records for approximately 15,000 entries. It is intended for research and analysis on physical activity, calorie expenditure, biometric tracking, and health patterns.
+
+**Description**
+
+This dataset collates two sources:
+
+- **Exercise Data**: Contains user demographics and biometric measurements during exercise sessions.
+- **Calories Data**: Logs calorie burn totals per user corresponding to activities tracked in the exercise dataset.
+
+This unified dataset supports detailed epidemiological, machine learning, or personal tracking studies to correlate biometric and demographic features with calorie expenditure.
+
+**Columns & definitions**
+
+From Exercise Data (`raw_exercise.csv`):
+
+- `User_ID`: Unique numerical ID for each individual.
+- `Gender`: Participant's gender (e.g., 'male', 'female').
+- `Age`: Age of participant (years).
+- `Height`: Height in centimeters.
+- `Weight`: Weight in kilograms.
+- `Duration`: Duration of exercise (minutes).
+- `Heart_Rate`: Heart rate measured during or after exercise (beats per minute).
+- `Body_Temp`: Body temperature measured during or after exercise (degrees Celsius).
+
+From Calorie Burn Data (`raw_calories.csv`):
+- `User_ID`: Unique numerical ID (to match/merge with above).
+- `Calories`: Calories burned as recorded/calculated for each session.
+
+**Usage notes**
+
+- **Merging**: Merge by `User_ID` for a complete record per exercise session.
+- **Applications**: Useful for regression, classification, health analytics, personalized fitness algorithms, and educational demos.
+- **Preprocessing**: Check for duplicates/missing values when merging. Normalize units if using for modeling.