Using Python for Machine Learning - notes

REGRESSION - Cities for predictor

Bonn
Budapest
Cork
Paris
Kiruna

CLASSIFICATION - optimal settings

KNN highest: .84375 (longitude, latitude; neighbors = 5)
Decision tree highest .78125 (longitude, latitude; split = 19)
Forest of trees highest .90625 (longitude, latitude; split = 2, trees = 15)
Bayes highest .84375 (latitude)
DNN highest so far .75 (latitude; layers = 5, epochs = 20, batchsize = 10, outputs = 64)