COE Coffee Aution Price Prediction

Course project of Statistical Learning and Deep Learning

Summary of content

• Web scrapped data from Alliance for Coffee Excellence to obtain the auction and competition list of 18 countries
• Applied text-mining techniques to analyze and categorize the coffee flavor characteristics described.
• Employed a diverse range of machine learning models, encompassing linear regression, lasso regression, ridge regression, random forest, and XGBoost and deep learning models, such as MLP, to refine the price prediction process.
• Found out the importance of flavor, aside from symbolic attributes, of coffee in affecting the auction price.

Countries we in vestigated

Distribution of target variable

Prediction result

Preference of different buyers toward different flavors