🛒 Retail Price Prediction Using Machine Learning

Predicting Canadian retail product prices (2017–2025) using:

• 📈 Linear Regression and 🌲 Random Forest Regression
• 🔍 Features: Year, Province, Product Category, Tax Rate, and more

---

✅ Key Accomplishments

Metric	Linear Regression	Random Forest
RMSE	1.24	1.36
R² Score	0.94	0.93

---

🔍 Project Overview

This project analyzes historical Canadian retail pricing data (2017–2025) to estimate pre-tax product prices. It applies supervised learning models and exploratory analysis to uncover how pricing varies by product category, location, time, and tax structure. Built to demonstrate end-to-end applied machine learning for real-world decision support.

---

📊 EDA Highlights

• 🍼 Top Expensive Items: Infant formula ($32), salmon, and beef rib/striploin cuts
• 🧾 Tax Variation: Provinces 2–6 showed the highest average tax (~15%), while Province 10 had the lowest (~5%)
• 📈 Price Trend: Average price rose steadily from $4.72 (2017) to $6.44 (2025)
• 📦 Category Spread: ‘Meat & Poultry’ and ‘Infant & Baby Products’ had the highest price variability

📁 Visualizations are saved in the plots/ folder.

---

🤖 ML Workflow

• Target: VALUE (pre-tax retail price)
• Features Used: Year, Month, GEO, Product Category, Product Name, Taxable, Essential, Total Tax Rate
• Preprocessing: One-hot encoding of categorical features
• Models Applied:
  • LinearRegression() from sklearn
  • RandomForestRegressor() from sklearn
• Evaluation Metrics: RMSE and R² Score

📌 See visual results in plots/PredictedVSActual.png

---

🗂️ Project Structure

retail-price-prediction/
│
├── Retail_Project.ipynb               # Main Jupyter notebook
├── readme.md                          # Project summary (this file)
├── requirements.txt                   # Python dependencies
├── summary.txt                        # Short findings summary
│
├── data/
│   ├── Retail_Prices_of_Products.csv  # Full dataset (cleaned)
│   └── Data_Dictionary.txt            # Column descriptions
│
├── plots/
│   ├── boxplot_by_category.png
│   ├── tax_by_province.png
│   ├── top10_expensive_products.png
│   ├── yearly_trend.png
│   └── PredictedVSActual.png

---

🛠️ How to Run

pip install -r requirements.txt
jupyter notebook

Open Retail_Project.ipynb and run all cells to reproduce results.

---

👤 Author

Arun Acharya
📍 Ottawa, Canada
🎓 Data Analyst in Training @ Willis College
📘 Google ML Crash Course

---

🔖 Tags

machine-learning regression retail data-analysis portfolio-project python

---

📄 License

This project is licensed under the MIT License.