This paper focused on User Behavior Prediction with Data Mining Framework based on optimized pattern mining techniques in a telecom company. The above skeleton considers the User behavior patterns and predicts the method they might act in the prospect. The main aims of this research work on numerous clustering techniques used to implement portfolio analysis, and it alienated preceding customers based on socio-demographic features using various clustering methods and Neural Network algorithms used to predict the user behavior. This paper proposes a unique records mining protocol named Cluster-based Temporal Mobile Sequential Style Mine (CTMSP-Mine) to efficiently unearth the Cluster-based Temporal Mobile Sequential Trend (CTMSPs). Proposed a novel prediction strategy to predict the user's subsequent behaviors using the discovered CTMSPs. The major payments of the job are actually that our experts make a proposal certainly not merely a unique formula for mining CTMSPs but also two nonparametric approaches for boosting the expected accuracy of the mobile users' habits. The proposed CTMSPs provide information, including both user clusters and temporal relations. Finally, by a speculative assessment of various substitute ailments, the planned procedure is shown to deliver exceptional functionality in relation to recall, f-measure, and precision.

DBSCAN, Color Image Segmentation, Clustering, Classification, Fuzz Logic, Pattern Mining, Performance Metrics.