Trajectories of born infants in the United States

• 

  Title of Project

• 

  Colliniarities

• 

  Causes on IM

• 

  Unstable Trajectories

• 

  Unstable Trajectories yield outputs

• 

  Linear Regression

Inspiration

The United states is ranked 174th in the world for Infant Mortality. These issues are highlighted through undernutrition, acute and chronic morbidities, lack of access to care, poor social support and wealth of the family. Factors such as these contribute to unstable health trajectories and with rapid and repeating episodes of illness, interspersed with short periods of recovery, it ultimately leads to death or for the cycle to repeat once again. By using time series analysis and Neural Networks we aim to project and predict future trends.

What it does

1) We use Linear Regression to predict a new born's weight. 2) Cleaned the data to mitigate collinearities, and harmonize various forms of data 3) Used PCA analysis to create clusters 4) Created recursive feature elimination with 10-fold cross validation. these use extra trees classifier or a lasso model to intelligently rank most important features

How we built it

Through the combination of SAS, R, and Python we joined our Computer Science and Statistics skills.

Challenges we ran into

After Data Wrangling, it seemed to be insufficient to do the proper analysis on.

Accomplishments that we're proud of

What we learned

What's next for Trajectories of born infants in the United States

Built With

• python
• r
• sas