Predicting a continuous-valued attribute with h2o in R

时间:2017-04-10 01:15:15

标签: r machine-learning deep-learning regression h2o

I have some data that follows an unknown multidimensional nonlinear relationship, for example:

x1 <- seq(0, 10, 0.1)
x2 <- seq(5, 15, 0.1)
y1 <- sin(x1)+exp(x2**2)+rnorm(length(x1))
y2 <- log10(abs(x1*x2)+1)+rnorm(length(x1))

I want to learn how to transform (x1,x2) into (y1,y2).

How can I use the package h2o in R to achieve this?

1 个答案:

答案 0 :(得分:0)

H2O(以及大多数机器学习包)不支持多个响应变量。在上述情况中,您有两个响应变量y1y2