Predict house price

Purpose: Predict the price of a house.

Dataset : boston_housing. Including crime rate, local property tax and other data, a total of 506 data.

Development brief : Considering the data with a large difference in the value range, the data is normalized, and because the number of data is small, k-fold cross-validation is used.

Model :

Distance of targets in 500 epochs and its acc&val fig:

Improve it into 160 epochs to avoid overfitting :

Predict :

Detail :

https://colab.research.google.com/drive/16t_fKfSd5kwzonQuWP77rJLBkp0kywCt?usp=share_link