corrosion_resistance_using_N_S_F.ipynb

{"cells":[{"metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","trusted":true},"cell_type":"code","source":"#import all imp libraries \nfrom numpy import loadtxt\nimport numpy as np\nimport matplotlib.pyplot as plt\nimport pandas as pd\nfrom sklearn.datasets import make_regression\nfrom sklearn.preprocessing import StandardScaler\nimport sklearn.model_selection as model_selection\nfrom keras.models import Sequential\nfrom keras.layers import Dense\nfrom keras.optimizers import SGD\nimport os \nimport csv\nimport time\nprint(\"libraries imported sucessfully !!!\")","execution_count":706,"outputs":[{"output_type":"stream","text":"libraries imported sucessfully !!!\n","name":"stdout"}]},{"metadata":{"_uuid":"d629ff2d2480ee46fbb7e2d37f6b5fab8052498a","_cell_guid":"79c7e3d0-c299-4dcb-8224-4455121ee9b0","trusted":true},"cell_type":"code","source":"#just to see what is in the data set \nimport csv\nwith open('../input/corrosion_final_data.csv','rt')as file:\n    data = csv.reader(file)\n    for row in data:\n        print(row)","execution_count":707,"outputs":[{"output_type":"stream","text":"['-1', '-1', '-1', '-572']\n['-1', '-1', '-1', '-591']\n['1', '-1', '-1', '-577']\n['1', '-1', '-1', '-594']\n['-1', '1', '-1', '-575']\n['-1', '1', '-1', '-587']\n['1', '1', '-1', '-568']\n['1', '1', '-1', '-589']\n['-1', '-1', '1', '-569']\n['-1', '-1', '1', '-592']\n['1', '-1', '1', '-574']\n['1', '-1', '1', '-588']\n['-1', '1', '1', '-573']\n['-1', '1', '1', '-593']\n['1', '1', '1', '-576']\n['1', '1', '1', '-558']\n['0', '0', '0', '-554']\n['0', '0', '0', '-615']\n['-2', '0', '0', '-557']\n['2', '0', '0', '-610']\n['0', '-2', '0', '-551']\n['0', '2', '0', '-611']\n['0', '0', '-2', '-558']\n['0', '0', '2', '-618']\n['0', '0', '0', '-586']\n['0', '0', '0', '-590']\n['0', '0', '0', '-582']\n['0', '0', '0', '-581']\n['0', '0', '0', '-591']\n['0', '0', '0', '-588']\n['0', '0', '0', '-587']\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"#in the above dataset we see that there are 3 input variables (X) and 1 output variable Y=F(X)\n#Input variables are N,S,F and output variable corrosion_pit_otential/mv \n#load the dataset \ndataset = pd.read_csv('../input/corrosion_final_data.csv',delimiter = ',')\nprint(\"dataset imported sucessfully !\")","execution_count":708,"outputs":[{"output_type":"stream","text":"dataset imported sucessfully !\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"print(dataset.iloc[:, :-1])","execution_count":709,"outputs":[{"output_type":"stream","text":"    -1  -1.1  -1.2\n0   -1    -1    -1\n1    1    -1    -1\n2    1    -1    -1\n3   -1     1    -1\n4   -1     1    -1\n5    1     1    -1\n6    1     1    -1\n7   -1    -1     1\n8   -1    -1     1\n9    1    -1     1\n10   1    -1     1\n11  -1     1     1\n12  -1     1     1\n13   1     1     1\n14   1     1     1\n15   0     0     0\n16   0     0     0\n17  -2     0     0\n18   2     0     0\n19   0    -2     0\n20   0     2     0\n21   0     0    -2\n22   0     0     2\n23   0     0     0\n24   0     0     0\n25   0     0     0\n26   0     0     0\n27   0     0     0\n28   0     0     0\n29   0     0     0\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"print(dataset.iloc[:, -1])","execution_count":710,"outputs":[{"output_type":"stream","text":"0    -591\n1    -577\n2    -594\n3    -575\n4    -587\n5    -568\n6    -589\n7    -569\n8    -592\n9    -574\n10   -588\n11   -573\n12   -593\n13   -576\n14   -558\n15   -554\n16   -615\n17   -557\n18   -610\n19   -551\n20   -611\n21   -558\n22   -618\n23   -586\n24   -590\n25   -582\n26   -581\n27   -591\n28   -588\n29   -587\nName: -572, dtype: int64\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"#Split datasets in input values(x) and output values(y)\n#data pre-processing for regression\nX = dataset.iloc[:, :-1].values\nY = dataset.iloc[:, -1].values\nprint(\"dataset split sucessfully !!\")","execution_count":711,"outputs":[{"output_type":"stream","text":"dataset split sucessfully !!\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"#standardize dataset and reshape the dataset\nX = StandardScaler().fit_transform(X)\nY = StandardScaler().fit_transform(Y.reshape(len(Y),1))[:, 0]\nprint(\"dataset reshaped sucessfully !!!!\")","execution_count":712,"outputs":[{"output_type":"stream","text":"dataset reshaped sucessfully !!!!\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"#split into train and test\ntrainX, testX, trainY, testY = model_selection.train_test_split(X, Y, train_size = 0.75, test_size = 0.25)\nprint(\"dataset splitted to training and testing data sucessfully !!!\")","execution_count":713,"outputs":[{"output_type":"stream","text":"dataset splitted to training and testing data sucessfully !!!\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"#Nh=Ns/(α∗(Ni+No))\n#ie; 30/(2(3+1)) = 3.75 approx 4\n#the first hidden layer has 4 nodes and uses relu\n#the second hidden layer has 3 nodes and uses relu activation function\n#the output layer had one node and uses sigmoid activation function \n#setting input layer with right no. of input features by setting it to 3 for 3 input features\nmodel = Sequential()\nmodel.add(Dense(4, input_dim=3, activation = 'relu', kernel_initializer= 'he_uniform'))\nmodel.add(Dense(3, activation='relu'))\nmodel.add(Dense(1, activation = 'sigmoid'))\nprint(\"specified sigmoid & activation functions for i/p & o/p layers !!!!\")","execution_count":714,"outputs":[{"output_type":"stream","text":"specified sigmoid & activation functions for i/p & o/p layers !!!!\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"#model will be fit with stochastic gradient descent with a learning rate of 0.01 and momentum of 0.9\nopt = SGD(lr = 0.01, momentum=0.9)\nmodel.compile(loss='mean_squared_error', optimizer='adam', metrics =['mean_squared_error'])\nprint(\"loss function specified sucessfully !!!!\")","execution_count":715,"outputs":[{"output_type":"stream","text":"loss function specified sucessfully !!!!\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"#training will be performed for 100 epochs and test set will be evaluated at the end of each epoch \nhistory = model.fit(trainX, trainY, validation_data=(testX, testY), epochs=100, batch_size =10)\nprint(\"fitted model sucessfully !!!\")","execution_count":716,"outputs":[{"output_type":"stream","text":"Train on 22 samples, validate on 8 samples\nEpoch 1/100\n22/22 [==============================] - 1s 31ms/step - loss: 1.2954 - mean_squared_error: 1.2954 - val_loss: 1.1270 - val_mean_squared_error: 1.1270\nEpoch 2/100\n22/22 [==============================] - 0s 609us/step - loss: 1.2920 - mean_squared_error: 1.2920 - val_loss: 1.1262 - val_mean_squared_error: 1.1262\nEpoch 3/100\n22/22 [==============================] - 0s 665us/step - loss: 1.2898 - mean_squared_error: 1.2898 - val_loss: 1.1254 - val_mean_squared_error: 1.1254\nEpoch 4/100\n22/22 [==============================] - 0s 708us/step - loss: 1.2871 - mean_squared_error: 1.2871 - val_loss: 1.1246 - val_mean_squared_error: 1.1246\nEpoch 5/100\n22/22 [==============================] - 0s 609us/step - loss: 1.2849 - mean_squared_error: 1.2849 - val_loss: 1.1240 - val_mean_squared_error: 1.1240\nEpoch 6/100\n22/22 [==============================] - 0s 618us/step - loss: 1.2829 - mean_squared_error: 1.2829 - val_loss: 1.1234 - val_mean_squared_error: 1.1234\nEpoch 7/100\n22/22 [==============================] - 0s 658us/step - loss: 1.2805 - mean_squared_error: 1.2805 - val_loss: 1.1228 - val_mean_squared_error: 1.1228\nEpoch 8/100\n22/22 [==============================] - 0s 659us/step - loss: 1.2785 - mean_squared_error: 1.2785 - val_loss: 1.1222 - val_mean_squared_error: 1.1222\nEpoch 9/100\n22/22 [==============================] - 0s 625us/step - loss: 1.2766 - mean_squared_error: 1.2766 - val_loss: 1.1215 - val_mean_squared_error: 1.1215\nEpoch 10/100\n22/22 [==============================] - 0s 678us/step - loss: 1.2750 - mean_squared_error: 1.2750 - val_loss: 1.1208 - val_mean_squared_error: 1.1208\nEpoch 11/100\n22/22 [==============================] - 0s 640us/step - loss: 1.2728 - mean_squared_error: 1.2728 - val_loss: 1.1202 - val_mean_squared_error: 1.1202\nEpoch 12/100\n22/22 [==============================] - 0s 631us/step - loss: 1.2707 - mean_squared_error: 1.2707 - val_loss: 1.1196 - val_mean_squared_error: 1.1196\nEpoch 13/100\n22/22 [==============================] - 0s 605us/step - loss: 1.2690 - mean_squared_error: 1.2690 - val_loss: 1.1190 - val_mean_squared_error: 1.1190\nEpoch 14/100\n22/22 [==============================] - 0s 664us/step - loss: 1.2670 - mean_squared_error: 1.2670 - val_loss: 1.1185 - val_mean_squared_error: 1.1185\nEpoch 15/100\n22/22 [==============================] - 0s 679us/step - loss: 1.2650 - mean_squared_error: 1.2650 - val_loss: 1.1181 - val_mean_squared_error: 1.1181\nEpoch 16/100\n22/22 [==============================] - 0s 592us/step - loss: 1.2630 - mean_squared_error: 1.2630 - val_loss: 1.1176 - val_mean_squared_error: 1.1176\nEpoch 17/100\n22/22 [==============================] - 0s 579us/step - loss: 1.2610 - mean_squared_error: 1.2610 - val_loss: 1.1169 - val_mean_squared_error: 1.1169\nEpoch 18/100\n22/22 [==============================] - 0s 621us/step - loss: 1.2583 - mean_squared_error: 1.2583 - val_loss: 1.1162 - val_mean_squared_error: 1.1162\nEpoch 19/100\n22/22 [==============================] - 0s 634us/step - loss: 1.2565 - mean_squared_error: 1.2565 - val_loss: 1.1156 - val_mean_squared_error: 1.1156\nEpoch 20/100\n22/22 [==============================] - 0s 637us/step - loss: 1.2546 - mean_squared_error: 1.2546 - val_loss: 1.1153 - val_mean_squared_error: 1.1153\nEpoch 21/100\n22/22 [==============================] - 0s 684us/step - loss: 1.2522 - mean_squared_error: 1.2522 - val_loss: 1.1150 - val_mean_squared_error: 1.1150\nEpoch 22/100\n22/22 [==============================] - 0s 1ms/step - loss: 1.2501 - mean_squared_error: 1.2501 - val_loss: 1.1145 - val_mean_squared_error: 1.1145\nEpoch 23/100\n22/22 [==============================] - 0s 671us/step - loss: 1.2486 - mean_squared_error: 1.2486 - val_loss: 1.1141 - val_mean_squared_error: 1.1141\nEpoch 24/100\n22/22 [==============================] - 0s 612us/step - loss: 1.2469 - mean_squared_error: 1.2469 - val_loss: 1.1135 - val_mean_squared_error: 1.1135\nEpoch 25/100\n22/22 [==============================] - 0s 687us/step - loss: 1.2449 - mean_squared_error: 1.2449 - val_loss: 1.1131 - val_mean_squared_error: 1.1131\nEpoch 26/100\n22/22 [==============================] - 0s 637us/step - loss: 1.2431 - mean_squared_error: 1.2431 - val_loss: 1.1126 - val_mean_squared_error: 1.1126\nEpoch 27/100\n22/22 [==============================] - 0s 631us/step - loss: 1.2416 - mean_squared_error: 1.2416 - val_loss: 1.1119 - val_mean_squared_error: 1.1119\nEpoch 28/100\n22/22 [==============================] - 0s 607us/step - loss: 1.2396 - mean_squared_error: 1.2396 - val_loss: 1.1113 - val_mean_squared_error: 1.1113\nEpoch 29/100\n22/22 [==============================] - 0s 605us/step - loss: 1.2375 - mean_squared_error: 1.2375 - val_loss: 1.1105 - val_mean_squared_error: 1.1105\nEpoch 30/100\n22/22 [==============================] - 0s 585us/step - loss: 1.2354 - mean_squared_error: 1.2354 - val_loss: 1.1098 - val_mean_squared_error: 1.1098\nEpoch 31/100\n22/22 [==============================] - 0s 602us/step - loss: 1.2335 - mean_squared_error: 1.2335 - val_loss: 1.1092 - val_mean_squared_error: 1.1092\nEpoch 32/100\n22/22 [==============================] - 0s 643us/step - loss: 1.2323 - mean_squared_error: 1.2323 - val_loss: 1.1085 - val_mean_squared_error: 1.1085\nEpoch 33/100\n22/22 [==============================] - 0s 718us/step - loss: 1.2302 - mean_squared_error: 1.2302 - val_loss: 1.1080 - val_mean_squared_error: 1.1080\nEpoch 34/100\n22/22 [==============================] - 0s 615us/step - loss: 1.2281 - mean_squared_error: 1.2281 - val_loss: 1.1074 - val_mean_squared_error: 1.1074\nEpoch 35/100\n22/22 [==============================] - 0s 625us/step - loss: 1.2260 - mean_squared_error: 1.2260 - val_loss: 1.1068 - val_mean_squared_error: 1.1068\nEpoch 36/100\n22/22 [==============================] - 0s 620us/step - loss: 1.2240 - mean_squared_error: 1.2240 - val_loss: 1.1064 - val_mean_squared_error: 1.1064\nEpoch 37/100\n22/22 [==============================] - 0s 613us/step - loss: 1.2224 - mean_squared_error: 1.2224 - val_loss: 1.1061 - val_mean_squared_error: 1.1061\nEpoch 38/100\n22/22 [==============================] - 0s 628us/step - loss: 1.2208 - mean_squared_error: 1.2208 - val_loss: 1.1056 - val_mean_squared_error: 1.1056\nEpoch 39/100\n22/22 [==============================] - 0s 642us/step - loss: 1.2187 - mean_squared_error: 1.2187 - val_loss: 1.1053 - val_mean_squared_error: 1.1053\nEpoch 40/100\n22/22 [==============================] - 0s 604us/step - loss: 1.2173 - mean_squared_error: 1.2173 - val_loss: 1.1052 - val_mean_squared_error: 1.1052\nEpoch 41/100\n22/22 [==============================] - 0s 589us/step - loss: 1.2156 - mean_squared_error: 1.2156 - val_loss: 1.1049 - val_mean_squared_error: 1.1049\nEpoch 42/100\n22/22 [==============================] - 0s 574us/step - loss: 1.2139 - mean_squared_error: 1.2139 - val_loss: 1.1046 - val_mean_squared_error: 1.1046\nEpoch 43/100\n22/22 [==============================] - 0s 608us/step - loss: 1.2128 - mean_squared_error: 1.2128 - val_loss: 1.1043 - val_mean_squared_error: 1.1043\nEpoch 44/100\n22/22 [==============================] - 0s 629us/step - loss: 1.2110 - mean_squared_error: 1.2110 - val_loss: 1.1042 - val_mean_squared_error: 1.1042\nEpoch 45/100\n22/22 [==============================] - 0s 627us/step - loss: 1.2099 - mean_squared_error: 1.2099 - val_loss: 1.1040 - val_mean_squared_error: 1.1040\nEpoch 46/100\n22/22 [==============================] - 0s 626us/step - loss: 1.2085 - mean_squared_error: 1.2085 - val_loss: 1.1036 - val_mean_squared_error: 1.1036\nEpoch 47/100\n22/22 [==============================] - 0s 629us/step - loss: 1.2067 - mean_squared_error: 1.2067 - val_loss: 1.1031 - val_mean_squared_error: 1.1031\nEpoch 48/100\n22/22 [==============================] - 0s 637us/step - loss: 1.2052 - mean_squared_error: 1.2052 - val_loss: 1.1028 - val_mean_squared_error: 1.1028\nEpoch 49/100\n22/22 [==============================] - 0s 649us/step - loss: 1.2036 - mean_squared_error: 1.2036 - val_loss: 1.1025 - val_mean_squared_error: 1.1025\nEpoch 50/100\n22/22 [==============================] - 0s 630us/step - loss: 1.2022 - mean_squared_error: 1.2022 - val_loss: 1.1021 - val_mean_squared_error: 1.1021\n","name":"stdout"},{"output_type":"stream","text":"Epoch 51/100\n22/22 [==============================] - 0s 643us/step - loss: 1.2004 - mean_squared_error: 1.2004 - val_loss: 1.1017 - val_mean_squared_error: 1.1017\nEpoch 52/100\n22/22 [==============================] - 0s 597us/step - loss: 1.1994 - mean_squared_error: 1.1994 - val_loss: 1.1012 - val_mean_squared_error: 1.1012\nEpoch 53/100\n22/22 [==============================] - 0s 654us/step - loss: 1.1975 - mean_squared_error: 1.1975 - val_loss: 1.1007 - val_mean_squared_error: 1.1007\nEpoch 54/100\n22/22 [==============================] - 0s 595us/step - loss: 1.1961 - mean_squared_error: 1.1961 - val_loss: 1.1001 - val_mean_squared_error: 1.1001\nEpoch 55/100\n22/22 [==============================] - 0s 690us/step - loss: 1.1946 - mean_squared_error: 1.1946 - val_loss: 1.0994 - val_mean_squared_error: 1.0994\nEpoch 56/100\n22/22 [==============================] - 0s 582us/step - loss: 1.1930 - mean_squared_error: 1.1930 - val_loss: 1.0986 - val_mean_squared_error: 1.0986\nEpoch 57/100\n22/22 [==============================] - 0s 604us/step - loss: 1.1911 - mean_squared_error: 1.1911 - val_loss: 1.0978 - val_mean_squared_error: 1.0978\nEpoch 58/100\n22/22 [==============================] - 0s 633us/step - loss: 1.1896 - mean_squared_error: 1.1896 - val_loss: 1.0971 - val_mean_squared_error: 1.0971\nEpoch 59/100\n22/22 [==============================] - 0s 566us/step - loss: 1.1880 - mean_squared_error: 1.1880 - val_loss: 1.0962 - val_mean_squared_error: 1.0962\nEpoch 60/100\n22/22 [==============================] - 0s 556us/step - loss: 1.1861 - mean_squared_error: 1.1861 - val_loss: 1.0954 - val_mean_squared_error: 1.0954\nEpoch 61/100\n22/22 [==============================] - 0s 596us/step - loss: 1.1851 - mean_squared_error: 1.1851 - val_loss: 1.0947 - val_mean_squared_error: 1.0947\nEpoch 62/100\n22/22 [==============================] - 0s 708us/step - loss: 1.1830 - mean_squared_error: 1.1830 - val_loss: 1.0941 - val_mean_squared_error: 1.0941\nEpoch 63/100\n22/22 [==============================] - 0s 772us/step - loss: 1.1816 - mean_squared_error: 1.1816 - val_loss: 1.0934 - val_mean_squared_error: 1.0934\nEpoch 64/100\n22/22 [==============================] - 0s 644us/step - loss: 1.1806 - mean_squared_error: 1.1806 - val_loss: 1.0927 - val_mean_squared_error: 1.0927\nEpoch 65/100\n22/22 [==============================] - 0s 687us/step - loss: 1.1792 - mean_squared_error: 1.1792 - val_loss: 1.0921 - val_mean_squared_error: 1.0921\nEpoch 66/100\n22/22 [==============================] - 0s 643us/step - loss: 1.1782 - mean_squared_error: 1.1782 - val_loss: 1.0915 - val_mean_squared_error: 1.0915\nEpoch 67/100\n22/22 [==============================] - 0s 644us/step - loss: 1.1771 - mean_squared_error: 1.1771 - val_loss: 1.0908 - val_mean_squared_error: 1.0908\nEpoch 68/100\n22/22 [==============================] - 0s 682us/step - loss: 1.1759 - mean_squared_error: 1.1759 - val_loss: 1.0901 - val_mean_squared_error: 1.0901\nEpoch 69/100\n22/22 [==============================] - 0s 665us/step - loss: 1.1744 - mean_squared_error: 1.1744 - val_loss: 1.0896 - val_mean_squared_error: 1.0896\nEpoch 70/100\n22/22 [==============================] - 0s 678us/step - loss: 1.1730 - mean_squared_error: 1.1730 - val_loss: 1.0890 - val_mean_squared_error: 1.0890\nEpoch 71/100\n22/22 [==============================] - 0s 690us/step - loss: 1.1717 - mean_squared_error: 1.1717 - val_loss: 1.0884 - val_mean_squared_error: 1.0884\nEpoch 72/100\n22/22 [==============================] - 0s 626us/step - loss: 1.1701 - mean_squared_error: 1.1701 - val_loss: 1.0878 - val_mean_squared_error: 1.0878\nEpoch 73/100\n22/22 [==============================] - 0s 605us/step - loss: 1.1687 - mean_squared_error: 1.1687 - val_loss: 1.0872 - val_mean_squared_error: 1.0872\nEpoch 74/100\n22/22 [==============================] - 0s 589us/step - loss: 1.1674 - mean_squared_error: 1.1674 - val_loss: 1.0866 - val_mean_squared_error: 1.0866\nEpoch 75/100\n22/22 [==============================] - 0s 629us/step - loss: 1.1662 - mean_squared_error: 1.1662 - val_loss: 1.0859 - val_mean_squared_error: 1.0859\nEpoch 76/100\n22/22 [==============================] - 0s 594us/step - loss: 1.1644 - mean_squared_error: 1.1644 - val_loss: 1.0853 - val_mean_squared_error: 1.0853\nEpoch 77/100\n22/22 [==============================] - 0s 576us/step - loss: 1.1629 - mean_squared_error: 1.1629 - val_loss: 1.0846 - val_mean_squared_error: 1.0846\nEpoch 78/100\n22/22 [==============================] - 0s 589us/step - loss: 1.1618 - mean_squared_error: 1.1618 - val_loss: 1.0839 - val_mean_squared_error: 1.0839\nEpoch 79/100\n22/22 [==============================] - 0s 598us/step - loss: 1.1606 - mean_squared_error: 1.1606 - val_loss: 1.0832 - val_mean_squared_error: 1.0832\nEpoch 80/100\n22/22 [==============================] - 0s 582us/step - loss: 1.1592 - mean_squared_error: 1.1592 - val_loss: 1.0827 - val_mean_squared_error: 1.0827\nEpoch 81/100\n22/22 [==============================] - 0s 606us/step - loss: 1.1580 - mean_squared_error: 1.1580 - val_loss: 1.0821 - val_mean_squared_error: 1.0821\nEpoch 82/100\n22/22 [==============================] - 0s 659us/step - loss: 1.1567 - mean_squared_error: 1.1567 - val_loss: 1.0815 - val_mean_squared_error: 1.0815\nEpoch 83/100\n22/22 [==============================] - 0s 649us/step - loss: 1.1551 - mean_squared_error: 1.1551 - val_loss: 1.0808 - val_mean_squared_error: 1.0808\nEpoch 84/100\n22/22 [==============================] - 0s 642us/step - loss: 1.1539 - mean_squared_error: 1.1539 - val_loss: 1.0799 - val_mean_squared_error: 1.0799\nEpoch 85/100\n22/22 [==============================] - 0s 671us/step - loss: 1.1524 - mean_squared_error: 1.1524 - val_loss: 1.0791 - val_mean_squared_error: 1.0791\nEpoch 86/100\n22/22 [==============================] - 0s 613us/step - loss: 1.1508 - mean_squared_error: 1.1508 - val_loss: 1.0784 - val_mean_squared_error: 1.0784\nEpoch 87/100\n22/22 [==============================] - 0s 635us/step - loss: 1.1496 - mean_squared_error: 1.1496 - val_loss: 1.0778 - val_mean_squared_error: 1.0778\nEpoch 88/100\n22/22 [==============================] - 0s 599us/step - loss: 1.1476 - mean_squared_error: 1.1476 - val_loss: 1.0772 - val_mean_squared_error: 1.0772\nEpoch 89/100\n22/22 [==============================] - 0s 632us/step - loss: 1.1464 - mean_squared_error: 1.1464 - val_loss: 1.0764 - val_mean_squared_error: 1.0764\nEpoch 90/100\n22/22 [==============================] - 0s 639us/step - loss: 1.1445 - mean_squared_error: 1.1445 - val_loss: 1.0758 - val_mean_squared_error: 1.0758\nEpoch 91/100\n22/22 [==============================] - 0s 642us/step - loss: 1.1433 - mean_squared_error: 1.1433 - val_loss: 1.0752 - val_mean_squared_error: 1.0752\nEpoch 92/100\n22/22 [==============================] - 0s 623us/step - loss: 1.1419 - mean_squared_error: 1.1419 - val_loss: 1.0746 - val_mean_squared_error: 1.0746\nEpoch 93/100\n22/22 [==============================] - 0s 596us/step - loss: 1.1403 - mean_squared_error: 1.1403 - val_loss: 1.0740 - val_mean_squared_error: 1.0740\nEpoch 94/100\n22/22 [==============================] - 0s 596us/step - loss: 1.1391 - mean_squared_error: 1.1391 - val_loss: 1.0732 - val_mean_squared_error: 1.0732\nEpoch 95/100\n22/22 [==============================] - 0s 613us/step - loss: 1.1379 - mean_squared_error: 1.1379 - val_loss: 1.0724 - val_mean_squared_error: 1.0724\nEpoch 96/100\n22/22 [==============================] - 0s 638us/step - loss: 1.1363 - mean_squared_error: 1.1363 - val_loss: 1.0716 - val_mean_squared_error: 1.0716\nEpoch 97/100\n22/22 [==============================] - 0s 648us/step - loss: 1.1349 - mean_squared_error: 1.1349 - val_loss: 1.0709 - val_mean_squared_error: 1.0709\nEpoch 98/100\n22/22 [==============================] - 0s 593us/step - loss: 1.1335 - mean_squared_error: 1.1335 - val_loss: 1.0704 - val_mean_squared_error: 1.0704\nEpoch 99/100\n22/22 [==============================] - 0s 617us/step - loss: 1.1323 - mean_squared_error: 1.1323 - val_loss: 1.0700 - val_mean_squared_error: 1.0700\nEpoch 100/100\n22/22 [==============================] - 0s 634us/step - loss: 1.1312 - mean_squared_error: 1.1312 - val_loss: 1.0696 - val_mean_squared_error: 1.0696\n","name":"stdout"},{"output_type":"stream","text":"fitted model sucessfully !!!\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"print(trainX)","execution_count":717,"outputs":[{"output_type":"stream","text":"[[-0.03809697 -0.03809697 -0.03809697]\n [-1.18100594  1.10481201 -1.18100594]\n [ 2.24772099 -0.03809697 -0.03809697]\n [-1.18100594 -1.18100594  1.10481201]\n [ 1.10481201 -1.18100594  1.10481201]\n [-0.03809697 -0.03809697 -0.03809697]\n [-0.03809697 -0.03809697 -0.03809697]\n [-2.32391492 -0.03809697 -0.03809697]\n [-0.03809697 -0.03809697 -0.03809697]\n [-1.18100594  1.10481201 -1.18100594]\n [ 1.10481201 -1.18100594  1.10481201]\n [ 1.10481201 -1.18100594 -1.18100594]\n [ 1.10481201  1.10481201  1.10481201]\n [ 1.10481201  1.10481201  1.10481201]\n [ 1.10481201  1.10481201 -1.18100594]\n [ 1.10481201 -1.18100594 -1.18100594]\n [-0.03809697 -0.03809697 -0.03809697]\n [-0.03809697 -0.03809697 -2.32391492]\n [-0.03809697 -0.03809697  2.24772099]\n [-0.03809697 -2.32391492 -0.03809697]\n [-0.03809697 -0.03809697 -0.03809697]\n [-0.03809697 -0.03809697 -0.03809697]]\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"print(trainY)","execution_count":718,"outputs":[{"output_type":"stream","text":"[-0.24559652  0.45058259 -1.57993981 -0.53567115  0.50859752  0.10249304\n -0.4196413   1.49485125 -0.30361145 -0.24559652 -0.30361145  0.33455274\n  0.39256766  1.43683633  0.85668707 -0.651701    0.04447811  1.43683633\n -2.04405922  1.84294081 -1.87001444 -0.47765622]\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"print(testX)","execution_count":719,"outputs":[{"output_type":"stream","text":"[[-0.03809697  2.24772099 -0.03809697]\n [-1.18100594  1.10481201  1.10481201]\n [-0.03809697 -0.03809697 -0.03809697]\n [-0.03809697 -0.03809697 -0.03809697]\n [-1.18100594 -1.18100594 -1.18100594]\n [ 1.10481201  1.10481201 -1.18100594]\n [-1.18100594  1.10481201  1.10481201]\n [-1.18100594 -1.18100594  1.10481201]]\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"print(testY)","execution_count":720,"outputs":[{"output_type":"stream","text":"[-1.63795474  0.56661244 -0.18758159  1.66889603 -0.47765622 -0.36162637\n -0.59368607  0.79867215]\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"#evaluating the model \n# training and testing for mean squared error\ntrain_mse = model.evaluate(trainX, trainY)\ntest_mse = model.evaluate(testX, testY)\nprint('Train: %.3f, Test: %.3f' % (train_mse[0], test_mse[0]))","execution_count":721,"outputs":[{"output_type":"stream","text":"22/22 [==============================] - 0s 91us/step\n8/8 [==============================] - 0s 391us/step\nTrain: 1.130, Test: 1.070\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"#plot loss during training\nplt.title('Loss ')\nplt.plot(history.history['loss'], label='train')\nplt.plot(history.history['val_loss'], label='test')\nplt.xlabel('epoch')\nplt.ylabel('loss')\nplt.legend()\nplt.show()","execution_count":722,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 1 Axes>","image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"metadata":{"trusted":true},"cell_type":"code","source":"#plot mean squared error during training\nplt.title('Mean Squared Error ')\nplt.plot(history.history['mean_squared_error'], label='train')\nplt.plot(history.history['val_mean_squared_error'], label='test')\nplt.xlabel('epoch')\nplt.ylabel('loss')\nplt.legend()\nplt.show()","execution_count":723,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 1 Axes>","image/png":"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\n"},"metadata":{"needs_background":"light"}}]}],"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat":4,"nbformat_minor":1}