our team name
As the world is quickly changing, so are the epidemics that affect the people who live in it. With that in mind, we have developed a program that can adapt and learn as times change. It can respond to new threats, and, given the time and computational power to learn, there are no pandemics it cannot prevent.
What it does
It integrates a neural network into a client side web page where users can enter in symptoms and get output of potential causes for their sickness.
How we built it
We created a multilevel neural network using large scale matrices. We also added in an extra function to the network to maximize its efficiency. The neural network was trained with online data and user experience. It was then integrated into a web format by first designing an HTML page and later converting that page into PHP to run the python code.
Challenges we ran into
Pythons library numpy which we were using to do lots of complex math, can not be run on the web. We did get a server working and got python running on it, but were unable to get Pythons library numpy working, therefor we opted to use an apache server to simulate the web sight.
Accomplishments that we're proud of
The minimalistic design of the webpage. Also the fact that the neural network is able to learn and improve its results over time.
What we learned
What's next for Sloth Analytics
The next step is to write all the numpy functions ourselves so that we are able to get the program running online where other users can interact with it. From there, we will make it so that doctors can input data and train the network, rather than relying on potentially unreliable users. We will also integrate the app with multiple platforms. We also would like to track the diseases and illnesses and monitor the globe for pandemics. It will also eventually be able to tell users what steps they can take to alleviate their symptoms (suggested medicine, scheduling doctors appointments, etc...)