This repository contains a machine learning project that predicts transport cost based on shipment and delivery features. The main goal is to build and evaluate regression models that can estimate the cost accurately.
| File / Folder | Description |
|---|---|
| Model_training.ipynb | Jupyter Notebook containing data preprocessing, model training, and evaluation. |
| ML_A1_Report.pdf | Final report summarizing the project, methods, and results. |
| Predictions/ | Folder for storing model predictions or output files. |
| Data/ | Folder for storing given and processed data. |
| .gitignore | Specifies files and folders ignored by Git. |
| README.md | Project documentation (this file). |
- Goal: Predict transport cost using regression models.
- Approach: Clean the dataset, handle missing values, transform features, and apply multiple regression algorithms.
- 9 Total models have been tested and evaluated: Adaboost, Decision Trees, Random Forest, Gradient Boosting, Linear Regression, Elastic net, Bayesian Ridge, Lasso Regression, K-nearest neighbours.
- Evaluation Metrics: R² score and RMSE.
- Python
- Pandas, NumPy, Scikit-learn
- Matplotlib, Seaborn
- Jupyter Notebook
- Open
Model_training.ipynbin Jupyter Notebook. - Run all cells step-by-step to train and evaluate models.
- According to the model u want to test, comment/uncomment the model and definition and its parameters in the jupyter code.
- Generated predictions will be saved as "submission.csv" (Rename it as you like).
Satyaram Mangena
For academic submission — Machine Learning Assignment 1.