Nji237/Nigeria-Political-Violence-Violence-Escalation-Analysis

Conducted data exploration on the Nigeria_ACLED.csv dataset, focusing on identifying factors associated with high ACLED fatalities. Explored variables such as actors, geographic locations, event types, and ACLED notes, using statistical techniques, data cleaning, and visualization in Python. Created insightful visualizations of key factors influencing fatalities, including bar charts and correlation plots. Applied machine learning techniques to the violence.csv dataset, using logistic regression and random forest models to predict the escalation of violence based on the binary 'Class' column. Selected appropriate evaluation metrics and compared model performances, improving the random forest model's accuracy by optimizing hyperparameters. Presented results clearly, demonstrating the impact of various features on violence escalation through visualizations and detailed explanations.