The rapid expansion of online social media platforms has revolutionized communication and connectivity, but it has also paved the way for the proliferation of fake accounts, posing serious threats such as the dissemination of false information and online fraud. Traditional methods for identifying fake accounts have proven inadequate in the face of evolving tactics employed by malicious actors. In response to this challenge, this paper proposes a novel approach leveraging machine learning and data science techniques to accurately detect fake accounts on social media platforms. The proposed methodology centers around the use of a gradient boosting algorithm combined with decision trees, which offers improved accuracy and robustness compared to conventional methods. Key attributes such as spam commenting, artificial activity, and engagement rate are utilized as input features to the algorithm, enabling the system to effectively distinguish between genuine and fake accounts. By integrating web scraping techniques, relevant data is extracted from social media profiles, facilitating the analysis of user behavior and activity patterns. One of the distinctive features of the proposed system is its ability to handle missing data effectively, ensuring reliable performance even in scenarios where complete information is not available. Moreover, the system's architecture is designed to be scalable and adaptable, allowing for seamless deployment across various social media platforms. Evaluation results demonstrate the efficacy of the proposed approach, with the system achieving high levels of accuracy in detecting fake accounts across different datasets. Furthermore, the system's real-time monitoring capabilities enable timely identification of fraudulent activities, empowering users and platform administrators to take proactive measures to mitigate risks. In conclusion, the proposed approach offers a promising solution to the persistent challenge of fake account detection on social media platforms. By leveraging advanced machine learning and data science techniques, the system provides a robust defense against fraudulent activities, thereby contributing to the preservation of trust and integrity in online communities.