I developed a predictive model using Random Forest to identify students at risk of social media addiction, using data from 705 records. I achieved 99.3% accuracy and an F1 score close to 1.0, validated with cross-validation (98.4% average accuracy). I used Python (pandas, matplotlib, seaborn, sklearn) for data cleaning, analysis, and visualization, identifying key variables such as sleep, mental health, age, and academic impact.

I identified key predictor variables (such as hours of sleep, self-assessed mental health, age, academic impact, and most used platform) that allow classifying students at risk of social media addiction, supporting the design of preventive strategies in mental health and education.

I analyzed over 73,000 ICFES Saber 11 test results from Antioquia (2017) to identify factors associated with academic performance. I developed a linear regression model in Python to predict critical reading scores, finding that socioeconomic status, access to books, the internet, and a computer at home significantly influence the results. I visualized the findings in an interactive dashboard in Power BI that highlights the relationship between family context.

I identified socioeconomic and family environment variables (such as number of books at home, internet and computer access, and social stratum) that significantly impact the ICFES critical reading score, providing useful evidence for educational policies focused on equity and performance.

Automation of reconciliations and accounting processes with Excel, support in financial reports, organization and digitization of accounting information, customer service and balance analysis.

Automation of reconciliations and accounting processes with Excel. Generation of financial reports and digitization of accounting information.