The Empirical Study of Applying Neuro to Escalate Marketing Performance

We develop a F-measure performance evaluation system by combining with RFM criteria and Bayesian classification, decision tree induction, Gini index, Neuro and by analyzing the factors with recall related and precision. By using this system, more profitable customers can be discovered and less unprofitable customers will be missed. The research collects and analyzes references on promoting buying rate; moreover, it introduces R-F-M (recency, frequency, monetary amount) criteria, brings up the idea of identify each individual customer to promote both the marketing profit and the customer´s lifetime value. The result shows that marketing performance derive from Neuro-weighted RFM model has the advantage over traditional RFM model by 27.80%.