Summary: | 碩士 === 銘傳大學 === 資訊工程學系碩士班 === 106 === Data Mining is a technology that explores the information that can predict the likely future occurrence and may be useful by studying the massive data in the past. This paper gets a real data set of customer coupons purchased from Kaggle website.The purpose of this paper is to find the customers who are interested in the coupons and to further solve the experimental difficulties. We use the k-nearest neighbor algorithm (KNN) to solve this problem. It predicts the category of new data based on the categories of training data whose k features are closest to the new data. The main core of KNN is to analyze the characteristics related to the categories to calculate the similarity between the data. Therefore, the problem to be overcome in this paper is how to find out the characteristics of customers who will purchase coupons and how to calculate the similarities between coupons and customers, so as to predict the first k customers who are most interested in the coupons, sort customers by interest level. Knowing which customers are interested in the products they can use marketing strategy for those who are interested, not only saving a lot of labor and advertising costs, but also not disturbing customers who are not interested in the product.
|