Are You Gonna Die?

Inspiration

Seeing the challenge that hospitals face, this prompt appealed to our group because we felt that we can help make a difference in the functionality of an institution's workflow.

What it does

Our model takes in large amounts of patient data including diseases, blood readings, socio-economics, etc. and it analyzes the content. The model is trained to find recurrences and trends, thus giving effective predictions of the likelihood of life in the patient's future.

How we built it

Using common data science packages, we trained a model in Python to read file from a large excel spreadsheet, and we analyzed the data. After running and reading through the data, statistics were calculated, and factors could be created to see the likelihood of life. To start with, we went through the data and cleaned out any data points that were missing values as the PCA was sensitive to NaN values. We also made sure to find outliers within datasets and determined whether or not certain values made sense in the context of what they were attributed. For instance, some age values in data were impossibly high so we took them out. We changed some value to be more easily read by the model, changing them into binary "yes or no" values if deemed necessary. All of these changes to the data were implemented through the use of loops to iterate through the rows under a specific column based on conditional checks.

Challenges we ran into

We ran into problems with understanding the patient data since none of the data had measurements included or descriptions. This made analysis exceptionally harder, but we researched each topic to better understand each reading.

Accomplishments that we're proud of

We are proud of the collaborative effort the team was able to put together in a short period of time. Each member became more fluent in health and hospital terminology, and we all demonstrated our adaptability to new projects.

What we learned

Through this project, we gained insight into machine learning and other data topics that could be used in future programs. Data and statistics were a massive topic that we became professionals in.