Cancer Diagnosis of malignant breast cancer with a Neural Network: This neural network implementation has a diagnosis accuracy of 98%.
Artificial Neural Networks (ANN) attempt in some degree to computationally model the inner workings of biological neural nets (like those in the brain.) Neural networks are good for problems where pattern recognition and data interpolation in noisy spaces form part of the problem. If sufficiently trained they have the ability to generalize beyond the data used for training - this is the definition of learning.
This code implements a multi-player synapse weighted neural network that is then trained on just under 700 real world doctors diagnoses.
It takes in various inputs such as tumor diameter, shape, density, etc and passes them through the trained network and yields a result of malignant or non-malignant.