File size: 1,490 Bytes
890bae1
 
3f03e68
 
dfc3d36
 
 
 
 
 
890bae1
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3f03e68
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
---
license: mit
language:
- en
pipeline_tag: tabular-classification
library_name: sklearn
tags:
- traffic
- sklearn
- tabular-classification
---
# TPnet-baseline

TPnet-baseline is a Random Forest classifier trained on smart mobility and traffic features to predict traffic congestion levels (Low, Medium, High) in urban environments.

## Model Details

- **Model type**: Random Forest Classifier
- **Input features**: 20 numerical features including vehicle count, road occupancy, weather, traffic light status, time-of-day, and more
- **Output**: Multiclass classification – `High`, `Medium`, `Low` traffic congestion
- **License**: MIT
- **Trained on**: Smart Mobility Traffic Dataset from Kaggle

## Training Details

- Train/test split: 80/20
- Accuracy (test): 99.9%
- F1 Score: 0.999
- Class-balanced via stratified sampling
- No overfitting observed

## Evaluation

| Metric     | Value  |
|------------|--------|
| Accuracy   | 99.9%  |
| F1 Score   | 0.999  |
| Model Size | ~1.2MB |

Confusion matrix and full report are available in the repository.

## How to Use

```python
import pickle

with open("traffic_predictor_rf.pkl", "rb") as f:
    model = pickle.load(f)

y_pred = model.predict(X_test)  # where X_test is a [n_samples, 20] array
```

## Limitations

- Does not account for live data
    
- Designed for offline batch inference
    
- Assumes all 20 features are properly preprocessed and scaled
    

## Authors

- Created by [@Yukin3](https://github.com/Yukin3)