The Prediction of Consumer Behavior from Social Media Activities

Area of Science:

• Computer Science
• Artificial Intelligence
• Machine Learning

Background:

• Advanced packing models complicate consumer behavior analysis.
• Traditional machine learning struggles with complex consumer behavior variants.

Purpose of the Study:

• Propose a novel deep learning ensemble model for enhanced consumer behavior classification.
• Improve the accuracy and efficiency of detecting consumer behavior variants.

Main Methods:

• Developed a new deep learning model utilizing an ensemble architecture.
• Incorporated two pre-trained learning algorithms in an optimized manner.
• The model followed four phases: data gathering, deep neural modeling, training, and evaluation.

Main Results:

• The ensemble model achieved 98.78% accuracy on the Facemg BIG-D15 database.
• Demonstrated superior performance compared to recent machine learning models.
• Outperformed existing models by over 8% in accuracy.

Conclusions:

• The proposed deep learning ensemble model effectively classifies consumer behavior variants.
• This approach offers a capable solution for detecting complex behavioral patterns.
• The model shows high precision and accuracy, advancing the field of consumer behavior analysis.