% Neural Network
% Created Wed Nov 08 01:33:00 2017
% Dragan Milicevic

close all
clear all
clc

% Set input data.
Inputs = [
65.86	74.43	78.89	98.16	66.22	162.95	155.28	63.90	69.61	75.32	89.78	63.54	155.99	148.14	83.17	83.17	58.90	151.35	76.92	145.82	57.47	141.36	68.89	123.87	110.30	112.80	212.39	221.32	230.24	228.45	232.02	232.02	249.87	208.82	214.18	223.10	221.32	223.10	223.10	240.95	208.82	214.18	210.61	215.96	212.39	233.81	210.61	208.82	207.04	205.25	212.39	199.90	207.04	224.88	223.10	274.86	276.64	242.73	260.58	265.94	215.96	217.75	264.15	265.94	233.81	257.01	255.94	214.18	260.58	264.15	226.67	249.87	241.84	260.58	223.10	244.52	229.53	240.95	221.85
0.038	0.075	0.150	0.300	0.600	1.200	1.500	0.038	0.075	0.150	0.300	0.600	1.200	1.500	0.150	0.300	0.600	1.200	1.300	1.500	0.600	1.200	1.300	1.500	0.600	1.500	0.075	0.150	0.300	0.600	1.200	1.500	1.800	0.075	0.150	0.300	0.600	1.200	1.500	1.800	0.150	0.300	0.600	1.200	1.500	1.800	0.300	0.600	1.200	1.500	1.800	1.500	1.800	0.075	0.150	0.300	0.600	1.200	1.500	1.800	0.075	0.150	0.300	0.600	1.200	1.500	1.800	0.150	0.300	0.600	1.200	1.500	1.800	0.600	1.200	1.500	1.800	1.500	1.800
4	4	4	4	4	4	4	16	16	16	16	16	16	16	36	36	36	36	36	36	64	64	64	64	100	100	4	4	4	4	4	4	4	16	16	16	16	16	16	16	36	36	36	36	36	36	64	64	64	64	64	100	100	4	4	4	4	4	4	4	16	16	16	16	16	16	16	36	36	36	36	36	36	64	64	64	64	100	100
]

% Set target data.
Targets = [
63.90	69.61	75.32	89.78	63.54	155.99	148.14	50.33	66.04	74.25	83.17	58.90	151.35	145.82	74.43	78.89	58.18	141.36	71.93	123.87	56.94	134.22	64.97	112.80	98.16	100.84	208.82	214.18	223.10	221.32	223.10	223.10	240.95	205.25	208.82	214.18	210.61	215.96	212.39	233.81	207.04	210.61	208.82	207.04	205.25	212.39	207.04	203.47	201.68	199.90	207.04	196.33	199.90	215.96	217.75	264.15	265.94	233.81	257.01	255.94	212.39	214.18	260.58	264.15	226.67	249.87	241.84	207.04	257.01	260.58	223.10	244.52	229.53	258.80	221.32	240.95	221.85	233.81	216.50
]

% Create a feed-forward backpropagation network with one hidden layer, 9 nodes in hidden layer,
% with input and output layer
net = feedforwardnet(9);

% Setup Division of Data for Training, Validation, Testing
net.divideFcn='dividerand'
net.divideMode = 'sample'
net.divideParam.trainRatio = 70/100;
net.divideParam.valRatio = 15/100;
net.divideParam.testRatio = 15/100;

% Set feed-forward backpropagation network train params
net.trainParam.show = 25
net.trainParam.epochs = 2000
net.trainParam.goal = 0
net.trainParam.min_grad = 1e-05
net.trainParam.max_fail = 2000
net.trainParam.mu = 0.0001
net.trainParam.mu_dec = 0.1
net.trainParam.mu_inc = 10
net.trainParam.mu_max = 10000000000
net.trainParam.lr = 0.25

% Choose Plot Functions
net.plotFcns = {'plotperform','plottrainstate','ploterrhist', 'plotregression'};

% Train the Network
net = train(net, Inputs, Targets)

% Test the Network
Outputs = net(Inputs)
Errors = gsubtract(Targets, Outputs);
Performance = perform(net, Targets, Outputs)