Support for Regression Problems

[cancan101]

Where the output is real values (R^n) and the loss function is something like MSE.

[apark263]

You can use the sum_squared cost (see neon/transforms/sum_squared.py)

We'll try to provide an example at some point, but it should be as simple
as swapping the CrossEntropy item in the CostLayer to be SumSquaredDiffs
instead

On Mon, May 18, 2015 at 3:41 PM, Alex Rothberg notifications@github.com
wrote:

Where the output is real values (R^n) and the loss function is something
like MSE.

—
Reply to this email directly or view it on GitHub
#33.