saikumardurgavajula/Algerian-forest

# Algerian Forest Fire Prediction & Deployment ## 📝 Project Overview This is an end-to-end machine learning project that predicts the **Fire Weather Index (FWI)** based on meteorological data from the Algerian Forest Fires dataset. The project covers the entire data science lifecycle, from data cleaning and exploratory data analysis to model training and deployment via a Flask web application. The core task is **regression**, where the final selected model—**Ridge Regression**—is deployed to predict the FWI, a key indicator of fire danger. --- ## Dataset Information The dataset contains 243 instances recorded between June and September 2012, covering two regions in Algeria: Bejaia and Sidi-Bel Abbes. **Key Attributes:** - **Meteorological Data:** Temperature, Relative Humidity (RH), Wind Speed (Ws), Rain. - **FWI System Components:** Fine Fuel Moisture Code (FFMC), Duff Moisture Code (DMC), Drought Code (DC), Initial Spread Index (ISI), and Buildup Index (BUI). - **Target Variable:** Fire Weather Index (FWI). --- ## ⚙️ Project Workflow This project is structured into three main phases: 1. **Data Analysis & Feature Engineering:** Cleaning the raw data, performing exploratory data analysis (EDA) to find insights, and preparing features for modeling. 2. **Model Training:** Training and evaluating several regression models to identify the best-performing one. 3. **Deployment:** Building a Flask web application to serve the trained model and allow for real-time predictions. --- ## 🛠️ Key Technical Steps ### 1. Data Cleaning and Preprocessing - Handled **missing values** and structural errors in the raw dataset. - Created a `Region` feature to distinguish between the "Bejaia" (0) and "Sidi-Bel Abbes" (1) regions. - Converted all feature data types to numerical (`int` and `float`) for analysis. - Encoded the categorical `Classes` feature (`fire`, `not fire`) into binary format (1, 0). ### 2. Exploratory Data Analysis (EDA) - Utilized histograms to examine …