KNN using tensorflow js

Predict housing price according to longitude and latitude

1. features --> longitude and latitude
2. labels --> house price
3. k --> number of neighboures considered
4. pp --> predictionPoint

1. get the distance from the pp ( s = (x**2 + y**2)**0.5 )
2. relate distance to house price
3. sort the tensors relative to the pp
4. get average of the top 'k' prices

[x] test with https://stephengrider.github.io/JSPlaygrounds/, dummy data [x] simple workflow on the playground [] migrate to vscode