Algerian Forest Fire

• EDA for Algerian Forest Fire dataset.
• Classification and Regression Models for predicting Classes between fire and not-fire and Temperature respectively.

ML Models

Regression

1. Linear
2. Lasso
3. Ridge
4. ElasticNet
5. SVR
6. Decision Tree
7. Random Forest
8. AdaBoost
9. XGBoost
10. CATBoost
11. Polynomial (degree 2)

Classification

1. Logistic
2. K-Nearest Neighbor
3. SVC
4. Naive Bayes
5. Decision Tree
6. Random Forest
7. AdaBoost
8. XGBoost
9. CATBoost

Deployed on Heroku

• Deployed Link: https://aff-prediction.herokuapp.com/
• Classification Prediction API
  • API: https://aff-prediction.herokuapp.com/classification
  • BODY: JSON data

  {
  	"day": 21,
  	"month": "Aug",
  	"Temperature": 36,
  	"RH": 58,
  	"Ws": 19,
  	"Rain": 0.0,
  	"FFMC": 88.6,
  	"DMC": 29.6,
  	"DC": 141.1,
  	"ISI": 9.2,
  	"Region": "Bejaia"
  }

• Regression Prediction API
  • API: https://aff-prediction.herokuapp.com/regression
  • BODY: JSON data

  {
  	"day": 21,
  	"month": "Aug",
  	"RH": 58,
  	"Ws": 19,
  	"Rain": 0.0,
  	"FFMC": 88.6,
  	"DMC": 29.6,
  	"DC": 141.1,
  	"ISI": 9.2,
  	"Classes": "fire",
  	"Region": "Bejaia"
  }