It was the eighth day of advanced mathematical and statistics class. Today’s lecture was on the topic Randomness, particularly in the context of spatial data analysis. Gary introduced two key types of randomness:

• Uniform Randomness:  Where points are distributed evenly across a space, with each location having an equal probability of selection.
• Poisson Randomness (Completely Spatial Randomness):  A distribution where events occur independently over space, following a Poisson process.

Additionally, he discussed Point Patterns and Point Processes, explaining how they help in modeling and analyzing spatial distributions of events. The lecture provided insights into how randomness plays a role in spatial statistics and its applications in real-world scenarios.

It was the seventh day of advanced mathematical and statistics class. Today, Gary introduced us to a new dataset which consists of Armed Conflict Location and Event Data (ACLED) on the regions India and USA.

We got a solid introduction to the dataset, discussing the variables involved, such as location, event type, date, and the involved parties. Then, we discussed on clustering methods, specifically K-means clustering. It’s particularly useful in our dataset because we can group regions based on the similarity of conflict types or intensity. We talked about choosing the number of clusters, the importance of centroids, and the iterative process of refining the clusters until they converge.

It was clear that applying clustering algorithms to the ACLED dataset could yield some valuable results.

The questions I have are:-

1. How might the nature of conflict (e.g., protest, violence, battles) affect the clustering results in different regions?
2. Can we use the clusters to predict future conflict hotspots or provide insights for conflict prevention?
3. How are the locations of conflicts distributed across different states or regions within India and the USA? Are certain areas more prone to conflict?
4. Are there any missing or incomplete data points in the dataset, and how should they be handled during analysis?

It was the sixth day of advanced mathematical and statistics class. Today, my groupmate Ahmed and I worked on analysing the fatal police shootings dataset. We cleaned and preprocessed the data, ensuring the correct classification of race, gender, and body camera usage. Together, we formulated key research questions and planned statistical comparisons and visualizations to highlight important patterns.

I focused on conducting statistical tests like chi-square and correlation analysis to identify significant relationships between variables, while Ahmed worked on descriptive statistics and population normalization. I also handled the Python scripting for statistical tests, whereas Ahmed wrote the data visualization scripts. We both debugged the code to ensure accuracy and reproducibility.

For the report, I wrote about mental health factors, circumstances of shootings, and body camera usage trends in the Findings section, while Ahmed covered racial disparities, age/gender analysis, and per capita rates. I also structured the Methodology (Appendix A) for clarity, and together, we worked on the Discussion section, summarizing key insights and policy recommendations.