We wanted to use Layered Neural Networks to give a more accurate prediction.
What it does
It calculates the probability that an individual will survive for 14 days after a stroke or stroke-like symptoms.
How we built it
We passed data from the International Stroke Trial into a 3-layer, supervised learning algorithm to train the AI.
Challenges we ran into
Small data set.
Accomplishments that we're proud of
The AI is highly realistic and can accurately predict about 90% of cases encountered.
What we learned
Using basic neural networks to train an AI to predict more accurately.
What's next for Assist
To incorporate more recent and larger data set. Also be able to use the setup neural network and expand to other datasets.