A Study on the Big Data Analysis of Retailing Product Strategy
2019
Hochschulschrift
Zugriff:
107
Recently years, Technology expand rapidly makes artificial intelligence rise quickly. Variety of data mining technologies are used widely throughout the online retail field, and combining series of well-know business index about customer availability and value. Data mining already become a common method, and an important tool for market advantage competition. This paper applies the structural of the RFM model through the online retail consumption data in the UCI database. And use K-Means clustering model to classify customers, apply Scikit-Learn cluster analysis K-Means model of Python software for data analysis. We divide the labeled data into training models and test model for construction and analysis model fit. According to the empirical analysis results of this paper, the classify results of clustering analysis have more difference in per group, and the customers who are more favorable to the enterprise (category 2) only include 26 customers among 8,082 customers. Therefore, the accuracy is lower to construct the decision tree model, and the overall accuracy is about 74%. Among them, the customers of category 2 are not classified correctly. The results explain the importance of collecting information, therefore increased the data collection, data pre-processing, and using multiple models for model fit analysis can improve results.
Titel: |
A Study on the Big Data Analysis of Retailing Product Strategy
|
---|---|
Autor/in / Beteiligte Person: | Lai,Jun-Ji ; 賴俊吉 |
Link: | |
Veröffentlichung: | 2019 |
Medientyp: | Hochschulschrift |
Sonstiges: |
|