This repository contains my complete solution for Assignment 1 of the Machine Learning Practices (MLP) course. The task was conducted as a Kaggle competition, where the objective is to predict flight ticket prices based on provided historical data.
In this assignment, we were given a training dataset and a test dataset. The labels for the test set were hidden, and we had to submit predicted prices for the test data on Kaggle. The assignment is evaluated via a leaderboard based on model performance.
- Deadline: July 2, 2025
- Platform: Kaggle Notebooks
- Kaggle Competition Link
| File Name | Description |
|---|---|
mlp-flight-price-prediction-assignment-1.ipynb |
Kaggle Notebook with full code, EDA, model building, and results |
README.md |
Project overview and submission details (this file) |
- Data Type Identification
- Descriptive Statistics of Numerical Columns
- Handling Missing Values
- Duplicate Removal
- Outlier Detection
- Data Visualization & Insights
- Feature Engineering
- Scaling & Encoding
A minimum of 7 different models were implemented:
- Linear Regression
- Decision Tree Regressor
- Random Forest Regressor
- Gradient Boosting
- K-Nearest Neighbors
- Support Vector Regression
- XGBoost
Hyperparameter tuning was performed on 3+ models using GridSearchCV and RandomizedSearchCV.
- Mean Absolute Error (MAE)
- Root Mean Squared Error (RMSE)
- R² Score
A brief video (under 10 min) explaining the structure, logic, and approach used in the notebook:
Watch on Google Drive
Notebook is hosted on Kaggle:
View on Kaggle
Dewang Gandhi
B.Tech Student | Machine Learning Enthusiast
GitHub: @dewanggandhi01
Email: [email protected] / [email protected]
This project is submitted as part of an academic assignment. All rights reserved.