Gladne/Tourism-price-prediction---Zindi

Tourism Price Prediction – Zindi is a machine learning project that predicts tourist expenditure in Tanzania using survey data from the National Bureau of Statistics. The notebook covers data cleaning, EDA, feature engineering, and model building to generate accurate spending predictions and support insights into tourism sector growth. # Tourism Price Prediction — Zindi Knowledge Challenge ## Project Overview This project predicts tourism costs using machine learning based on traveler behavior, activities, and package details. ## Objective Build a model to accurately predict total_cost. --- ## Dataset The dataset was obtained from the **Zindi Tourism Knowledge Challenge** and contains: - Demographic information - Travel group composition - Tourist experience factors - Target variable: `total_cost` Files included: - `train.csv` — Training dataset - `test.csv` — Test dataset - `SampleSubmission.csv` — Submission format - `VariableDefinitions` - Variable Descriptions --- ## Key Steps - Data cleaning and missing value handling - Feature engineering (group_size, region, total_nights) - Encoding (binary, one-hot, target encoding) - Log transformation of target variable --- ## Models used and Results | Model | RMSE | R² | |----------------|--------|------| | **XGBoost** | **4.05M** | **0.41** | | Random Forest | 4.68M | 0.26 | | Decision Tree | 5.91M | -0.17 | --- ## Conclusion XGBoost achieved the best performance, showing strong ability to capture complex patterns in tourism spending. ## How to Run ```bash git clone github.com cd Tourism-price-prediction---Zindi pip install -r requirements.txt ``` Run the notebook: ```Tourism_price_prediction.ipynb```