Neural Network type conversion problem

Question

So Im basically trying to make a neural network that tells if a defined set of points (vector2s) are in a line. The thing is that I want to end the output with a confidence of how "liny" it is and my first node is a list of vector2. How can I start with vector2 and end in a float? Or is there a logical way to convert a v2 to a float?

Answer 1

Answer by Kishotta · Sep 24, 2017 at 06:13 AM

While you could possibly do this with an ANN, you could also search for a C# Linear Algebra or Statistics library and use linear regressions to solve this with considerably less time and effort.

Add comment · Show 6 · Share

Guy_Yome · Sep 24, 2017 at 06:42 AM 0

Share

Well, I wanna see if its a line just as my first try at neural networks. Could have chosen to look like face or something else.

Kishotta Guy_Yome · Sep 24, 2017 at 06:51 AM 0

Share

If you've never made an ANN, you might look at MNIST. It's sorta the "Hello World" of neural nets.

Guy_Yome Kishotta · Sep 24, 2017 at 03:50 PM 0

Share

So in my case I could maybe normalize the vectors and use the distance between the most distant points as the value of a particular node. Right?

Show more comments

Neural Network type conversion problem

1 Reply

Your answer

Welcome to Unity Answers

Follow this Question

Related Questions