Tutorials
Below you can find tutorials or guides for various quantitative data techniques in machine learning, network analysis, and handling big data. These tutorials are products from multiple courses taken at Penn State.
Preferential Attachment: This tutorial provides a guide to popularity, or preferential attachment, in networks. Preferential attachment means that the more connected a node is, the more likely it is to receive new links. Nodes with higher degrees have a stronger ability to attract links added to the network.
Model Comparison and Hyperparameter Tuning: In this tutorial, I use electoral violence to fine-tuning algorithms and evaluate the performance of machine learning models. My data comes from the Varieties of Democracy (VDEM) dataset and Electoral Contention and Violence (ECAV) dataset. Using these two sources, the goal of my model is to predict electoral violence based on a handful of covariates: Gender Quota, Equal Protection Index, Regime Type, Electoral System, and actor/target types. I assess the performance of this model using different performance metrics, diagnose potential bias/variance issues with the model, and fine tune the parameters for optimizing the model.
Measurement and Reclassification of Data: In this tutorial, I review the levels of measurement for variables, how they should be used, and provide examples of each type in application. I also discuss reclassification issues, then provide tutorials on ordinal regression and Multiple Correspondence Analysis (MCA).