Inspiration
The Safe Haven_DataNauts project was inspired by the growing homelessness crisis in California. With significant disparities in service access and utilization across counties, the project aims to use data-driven insights to inform policies and improve resource allocation.
What it does
The project analyzes hospitalization and PIT count data to uncover trends in homelessness, hospital utilization, and system performance across California counties. It uses predictive models and clustering to identify key drivers of homelessness and provide actionable insights for policy interventions.
How we built it
We leveraged Python libraries such as Pandas, NumPy, Matplotlib, Seaborn, and Scikit-learn to clean, visualize, and model the data. The pipeline included data standardization, feature engineering, and the development of a linear regression model to predict homelessness variation.
Challenges we ran into
We faced challenges with missing data, inconsistent formats, and sparse demographic information, which required significant data cleaning and standardization. Additionally, some counties had unique local factors that the model couldn’t capture, leading to prediction deviations.
Accomplishments that we're proud of
We successfully cleaned and standardized a complex dataset, built interpretable models, and identified key homelessness trends. The clustering analysis also provided valuable insights into regional differences across California counties.
What we learned
We learned the importance of thorough data cleaning and the need for better demographic data. Additionally, we discovered the challenges of accounting for regional variations and the value of interpretability in models aimed at informing policy.
What's next for Safe Haven_DataNauts
Next, we aim to expand the dataset with more granular data, refine our models, and explore real-time data integration. The goal is to create a public dashboard to provide ongoing insights and support targeted interventions for homelessness in California.
Log in or sign up for Devpost to join the conversation.