https://avabbas.github.io/academic-cv/tags/behavioral-analysis/Behavioral AnalysisHugoBlox Kit (https://hugoblox.com)en-usMon, 20 Jul 2026 00:00:00 +0000https://avabbas.github.io/academic-cv/media/icon_hu_5ac22c9ce43dded7.pnghttps://avabbas.github.io/academic-cv/tags/behavioral-analysis/https://avabbas.github.io/academic-cv/projects/mashhad-social-study/Mon, 20 Jul 2026 00:00:00 +0000https://avabbas.github.io/academic-cv/projects/mashhad-social-study/<h2 id="overview">Overview</h2> <p>A data-driven computational study investigating the gap between <strong>romantic interest and actual approach behavior</strong> among young adults in Mashhad, Iran, a city with a unique blend of religious tradition and modern student culture. Using a self-collected dataset of <strong>1,000 participants</strong>, this project applies machine learning and statistical modeling to uncover hidden behavioral patterns, paradoxes, and gender asymmetries in human social initiation.</p> <hr> <h2 id="-focus">🔍 Focus</h2> <p>The study examines how fear of rejection, self-perceived attractiveness, religiosity, daily habits, and digital behaviour relate to whether young adults actually approach someone they are interested in, and how these patterns differ by gender.</p> <hr> <h2 id="-methodology">🧪 Methodology</h2> <ul> <li><strong>Dataset:</strong> 1,000 participants (balanced gender distribution), 30-item structured questionnaire</li> <li><strong>Data Cleaning:</strong> Pandas, NumPy — Persian text-to-numeric encoding pipeline</li> <li><strong>Statistical Testing:</strong> SciPy — Independent T-Tests across 29 features with Bootstrap confidence intervals</li> <li><strong>Machine Learning:</strong> Scikit-learn — Random Forest, Gradient Boosting, Logistic Regression, K-Means clustering, Isolation Forest, t-SNE</li> <li><strong>Deep Learning:</strong> PyTorch — 4-layer neural network classifier + Autoencoder for anomaly detection</li> <li><strong>Explainability:</strong> SHAP — Feature importance analysis revealing that <strong>Social Sacrifice Index</strong> (willingness to risk social prestige) is the strongest predictor of approach behavior</li> <li><strong>Visualization:</strong> Matplotlib, Seaborn — 12 publication-quality figures</li> </ul> <hr> <h2 id="-results--impact">📊 Results &amp; Impact</h2> <ul> <li>Built a conceptual model mapping psychological, economic, and spatial factors to social action</li> <li>Generated <strong>12 high-resolution figures</strong> suitable for academic publication</li> </ul> <hr> <h2 id="-technologies-used">🛠️ Technologies Used</h2> <p><code>Python</code> <code>Pandas</code> <code>NumPy</code> <code>SciPy</code> <code>Scikit-learn</code> <code>PyTorch</code> <code>SHAP</code> <code>Matplotlib</code> <code>Seaborn</code> <code>Google Colab</code></p> <hr> <h2 id="-personal-contributions">💡 Personal Contributions</h2> <ul> <li>Designed and executed the complete <strong>end-to-end data science pipeline</strong></li> <li>Engineered <strong>200+ features</strong> including pairwise interactions and polynomial transformations</li> <li>Implemented <strong>3 anomaly detection methods</strong> (Isolation Forest, LOF, Autoencoder) with consensus voting</li> <li>Applied <strong>fairness metrics</strong> to ensure the model is not gender-biased</li> <li>Created an automated interpretation engine producing <strong>100+ analytical insights</strong></li> </ul> <hr> <p><strong>Keywords:</strong> <code>Computational Social Science</code> <code>Human-Computer Interaction</code> <code>Behavioral Machine Learning</code> <code>Gender Analysis</code> <code>Socio-Religious Context</code> <code>Anomaly Detection</code> <code>Explainable AI</code></p>