Decoding Consumer Behavior: Predictive Analytics of Social Media Advertising Effectiveness
Pradeep Kandula
DOI: https://doi.org/10.55041/ijsrem32456
2024-05-01
INTERANTIONAL JOURNAL OF SCIENTIFIC RESEARCH IN ENGINEERING AND MANAGEMENT
Abstract:In the digital age, social media advertising has become an indispensable tool for marketing. This study investigates the effectiveness of social media advertisements in influencing consumer purchasing decisions. Using a dataset obtained from Kaggle, this analysis incorporates demographic and socio-economic factors—specifically, age, gender, and estimated salary—to predict consumer behavior towards product purchases after exposure to advertisements. Our dataset consists of attributes such as User ID, gender, age, salary, and purchase history. We employed several machine learning classification models including logistic regression, decision trees, random forests, SVM, Naive Bayes, KNN, and Perceptron to identify the most predictive factors and to model purchasing likelihood. Outliers were removed and missing data imputed through statistical methods to ensure data integrity. The highest predictive accuracy was achieved using a Random Forest model with 20 trees, which demonstrated a 95% accuracy rate. This model was integrated into a user-friendly interface developed with Gradio, facilitating real-time predictions. Our findings can assist companies in tailoring their marketing strategies to target demographics most likely to respond positively to advertisements, thereby enhancing the efficiency and effectiveness of their marketing campaigns on social platforms. Keywords:- Social Media Advertising, Consumer Behavior, Predictive Analytics, Machine Learning, Classification Algorithms, Random Forest, Marketing Strategy, Purchase Intentions, Data Preprocessing, Gradio Interface.