Flight Price Prediction and Analysis This project focuses on predicting flight prices using machine learning techniques and performing exploratory data analysis (EDA) to understand the dataset. Here's a breakdown of what's included:

1. Exploratory Data Analysis (EDA) Data Import and Inspection: Loaded dataset from Clean_Dataset.csv using pandas. Explored basic information such as shape, data types, and statistical summaries. Handling Missing Values: Visualized missing data patterns using missingno. Duplicate Check: Ensured data integrity by checking for duplicate entries based on a unique column. Visualizing Categorical Data: Utilized Plotly Express to create bar charts and donut charts for categorical variables to understand their distributions. Price Distribution and Analysis: Analyzed price distribution and its range. Mean Price Analysis by Categories: Investigated mean prices across categorical columns.

2. Data Preprocessing Encoding: Applied one-hot encoding to categorical columns (airline, source_city, destination_city, departure_time). Column Transformation: Converted class column to binary and stops column to numerical format. Feature Selection: Dropped unnecessary columns (Unnamed: 0, flight) and retained relevant features (duration, days_left, price).

3. Modeling Regression Modeling: Built a Random Forest Regressor model to predict flight prices. Cross-validation: Used K-Fold cross-validation for robust evaluation. Evaluation Metrics: Calculated metrics such as R-squared, Mean Absolute Error (MAE), Mean Squared Error (MSE), and Root Mean Squared Error (RMSE). Visualization of Predictions: Visualized predicted prices against actual prices using scatter plots.

4. Analysis and Interpretation MAE and Price Range Comparison: Analyzed MAE relative to the price range to assess model performance. Residual Analysis: Examined residuals to check model assumptions. Cross-validation Consistency: Ensured consistency through cross-validated MAE. Feature Importance: Identified top features influencing price predictions using Random Forest feature importances. Prediction on New Data: Created a function to preprocess new data and predict flight prices. This repository provides a comprehensive analysis of flight prices, from initial data exploration to predictive modeling, aimed at understanding and predicting flight costs effectively.