artificial intelligence - Training neural network for XOR in Ruby -

i trying train feedforward network work perform xor operations ruby library ai4r. however, when evaluate xor after training it. not getting correct output. has used library before , gotten learn xor operation.

i using 2 input neurons, 3 neurons in hidden layer, , 1 layer output, saw precomputed xor feed forward neural network before.

require "rubygems" require "ai4r"  # create network with:  #   2 inputs  #   1 hidden layer 3 neurons  #   1 outputs  net = ai4r::neuralnetwork::backpropagation.new([2, 3, 1])     example = [[0,0],[0,1],[1,0],[1,1]]  result = [[0],[1],[1],[0]]   # train network  400.times |i|    j = % result.length    puts net.train(example[j], result[j])  end   # use it: evaluate data trained network puts "evaluate 0,0: #{net.eval([0,0])}"  # =>  evaluate 0,0: 0.507531383375123 puts "evaluate 0,1: #{net.eval([0,1])}"  # =>  evaluate 0,1: 0.491957823618629 puts "evaluate 1,0: #{net.eval([1,0])}"  # =>  evaluate 1,0: 0.516413912471401 puts "evaluate 1,1: #{net.eval([1,1])}"  # =>  evaluate 1,1: 0.500197884691668 

ted

you haven't trained enough iterations. if change 400.times 8000.times you'll come closer (and closer still @ 20000.times).

at 20000.times, get

puts "evaluate 0,0: #{net.eval([0,0])}"  # =>  evaluate 0,0: 0.030879848321403 puts "evaluate 0,1: #{net.eval([0,1])}"  # =>  evaluate 0,1: 0.97105714994505 puts "evaluate 1,0: #{net.eval([1,0])}"  # =>  evaluate 1,0: 0.965055940880282 puts "evaluate 1,1: #{net.eval([1,1])}"  # =>  evaluate 1,1: 0.0268317078331645 

you can increase net.learning_rate (but not much).