ShashankKumar2160/Algerian_Forest_Fire_Exploratory_Analysis

# 🔥 Exploratory Data Analysis (EDA) of Algerian Forest Fires ## 📌 Project Overview This project performs **Exploratory Data Analysis (EDA)** on the **Algerian Forest Fires Dataset** from the UCI Machine Learning Repository. The dataset contains meteorological and fire-related data collected from two regions in Algeria to help analyze fire occurrences and environmental patterns. ## 📂 Dataset Overview - **Total Instances:** 244 - **Total Attributes:** 13 - **Regions Covered:** - **Bejaia Region (1st region)** - **Sidi Bel-abbes Region (2nd region)** - **Target Variable:** Fire occurrence (`fire` or `not fire`) ## 🏷️ Features | Feature | Description | |---------|-------------| | `Date` | Date of observation | | `Temperature (°C)` | Daily average temperature | | `RH (%)` | Relative humidity | | `Ws (km/h)` | Wind speed | | `Rain (mm)` | Total daily rainfall | | `FFMC` | Fine Fuel Moisture Code (fire risk indicator) | | `DMC` | Duff Moisture Code (moisture in medium fuels) | | `DC` | Drought Code (long-term moisture deficit) | | `ISI` | Initial Spread Index (fire spread potential) | | `BUI` | Buildup Index (total fuel availability) | | `FWI` | Fire Weather Index (comprehensive fire risk indicator) | | `Classes` | Fire occurrence (`fire` = 1, `not fire` = 0) | ## 🎯 Objectives - Understand **fire occurrence patterns** based on meteorological variables. - Identify **correlations** between weather conditions and fire spread. - Perform **data visualization** and statistical analysis. - Develop insights for **fire prevention and management**. ## 📊 EDA Steps 1. **Data Preprocessing** - Handling missing values - Encoding categorical data - Data type conversions 2. **Descriptive Statistics** - Summary statistics (mean, median, mode, etc.) - Checking for outliers 3. **Data Visualization** - Distribution plots (histograms, boxplots) - Correlation heatmaps - Time-series analysis 4. **Feature Engineering** - Creating new meaningful features - Removing redundant columns ## 🛠️ To …