Category: Uncategorized

Proj1 details discussed with professor and and some details included in project2 how me and my teammate contributing each other with the professor

—where I grouped U.S. states by the number of political protests, riots, and incidents of violence against civilians—I’ve now taken a crucial next step: adjusting for population size.

This idea came from a helpful suggestion by my professor, and the more I thought about it, the more it made sense. States like California or Texas naturally show higher event counts simply because they have larger populations. But that doesn’t necessarily mean they’re more protest-prone than smaller states. To make fair comparisons, I needed to account for that population difference.

So this week, I:

Gathered population data for each U.S. state from the Census Bureau.

Combined that with my event data from the ACLED dataset.

Calculated normalized event rates by finding the number of each event type per 100,000 people in each state.

Discussed on point patterns:

Point Patterns
These refer to sets of points that are distributed within a defined space — essentially, an observed dataset of points in that space. Based on their spatial arrangement, point patterns are generally categorized into:

Poisson Patterns:
Points are randomly and independently scattered across the space, following a uniform distribution.

Clustered Patterns:
Points appear to group or cluster together, often due to some underlying factor or mutual influence.

Point Processes
A point process is a mathematical framework used to model the random distribution of points in space and/or time. It helps us understand how observed point patterns might be generated probabilistically.

We also explored an idea that we’d like to dive deeper into — using the Armed Conflict Location and Event Data (ACLED). The core thought is to select a specific location and try to predict the likelihood of a violent political demonstration occurring there in the future. The prediction would be based on analyzing past events at that location and its nearby neighbors, using historical patterns to estimate the probability of future conflict.

 

UrbanPop & Police Shootings (0.39): Moderate correlation; urban areas may experience more
crime, leading to greater police presence and use of force, though rural areas also see notable
police shootings.
UrbanPop & Murder (Negative Correlation): Indicates that murder rates aren’t necessarily
higher in urban areas, suggesting other socioeconomic factors may play a role.

Correlation values range from -1 to 1:
• 1.0 (Red): Perfect positive correlation (two variables increase together).
• 0.0 (White): No correlation.
• -1.0 (Blue): Perfect negative correlation (one variable increases, the other decreases).
I build this to show high murder areas also have high assault rate and those are urbanized areas,
so this pushes for need of more police surveillance. But this might not be true so I have another
correlation matrix
1.Crime & Police Shootings:
Murder (0.41): Moderate positive correlation; higher murder rates may lead to increased police
shootings due to heightened law enforcement presence or violent encounters.
Assault (0.44): Slightly stronger correlation than murder; physical altercations often result in
more police interventions.
Rape (0.51): The strongest correlation, suggesting sexual violence cases may involve higher
police use of force, possibly due to violent confrontations or high-risk arrests.

 

the above are the basic stats which are driven out from the Dataset i merged and also

These 2 bar chats show us how crime rank is split in between different frequencies due to
different social and geospatial reasons, I used to highlight the point on police shootings.

Findings:
• Higher crime rate regions exhibit a statistically significant increase in police shooting
incidents.
• More population a county has it has more chance of having more shooing
• Eastern states have lower population so even with population density they are fewer
police shootings
• Extreme temperature conditions don’t properly correlate with increased instances of
police shootings.
• It’s more directly connected to population than temperature.
• Population density and demographic variables influence police encounters, with densely
populated areas showing higher incidences even within same states.
• Statistical outliers suggest other factors apart from population and GDP playing a factor
in some
• Main conclusion we can draw is most of these factors are interlinked. Like population,
crime, poverty and all which leads to more police action in those areas.

These are the findings that me and my teammate Nikhil Valaja Found out and need to work on these

Analysed the police shooting data which was the professor was given by and also trying to found out the other related data to merge them and make it whole dataset.

Later i found the crime data from kaggle and county data from gigasheet.com.

Need to work on these

In my first class of Advanced Mathematical Statistics (MTH-522) taught by Professor Dr. Gary Davis, and we covered a lot of interesting topics.

I learned how to set up a work journal using a WordPress site, which was a valuable skill. The professor presented us with a dataset about police shootings in the United States, sourced from The Washington Post.

We also discussed how to write a concise and effective report, often referred to as a “Punchline report.” Additionally, we explored the concept of Activity Theory, which involves connecting a spontaneous question to a specific objective.

Need to learn and explore more about the Dataset in Coming weeks

Welcome to UMassD WordPress. This is your first post. Edit or delete it, then start blogging!