diff --git a/venv/outputmodel1/Dataset2-2.h5 b/venv/outputmodel1/Dataset2-2.h5
index 1b18d20..239d746 100644
Binary files a/venv/outputmodel1/Dataset2-2.h5 and b/venv/outputmodel1/Dataset2-2.h5 differ
diff --git a/venv/src/IdentifyInvalidSignals.py b/venv/src/IdentifyInvalidSignals.py
new file mode 100644
index 0000000..fffe9f2
--- /dev/null
+++ b/venv/src/IdentifyInvalidSignals.py
@@ -0,0 +1,146 @@
+# pip install keras
+# pip install tensorflow
+import keras
+import numpy as np
+import scipy.io
+
+from keras.models import Sequential
+from keras.layers import Activation, Flatten
+from keras.layers import Convolution2D
+from keras.layers.core import Dense
+from keras.utils import np_utils
+import matplotlib.pyplot as plt
+import pywt
+from keras.models import load_model
+
+
+def spectrum(vector):
+    A = np.fft.fft(vector)
+    ps = np.abs(A) ** 2
+    ps = ps[:len(ps) // 2]
+    return ps
+
+
+def normalize(someList):
+    amin = np.min(someList)
+    amax = np.max(someList)
+    mean = np.mean(someList)
+    for i in range(0, len(someList)):
+        someList[i] = someList[i] / amax
+    return someList
+
+
+def removereadings(someList):
+    return someList[:150]
+
+
+def customeProcess(someList):
+    return removereadings(normalize(someList))
+
+
+file = "/Users/rmramesh/Downloads/EEG_BCI_MI_AllSub/SubC_6chan_2LF_s3.mat"
+data = scipy.io.loadmat(file)
+print(data.keys())
+
+xdata = data.get("EEGDATA")
+print(xdata.shape)
+xdata = xdata.transpose(2, 0, 1)
+print(xdata.shape)
+
+ydata = data.get("LABELS")
+print(ydata.shape)
+
+xshape = xdata.shape
+numberOfItems = xshape[0]
+numberOfElectrodes = xshape[1]
+numberOfReadings = xshape[2]
+
+# Array shape after processing
+nrap = int(numberOfReadings / 2)
+
+xprocessed = np.zeros([numberOfItems, 150, 150], dtype=float)
+
+for i in range(0, numberOfItems):
+    for j in range(0, numberOfElectrodes):
+        # xprocessed[i][j] = spectrum(xdata[i][j])
+        xprocessed[i][j] = customeProcess(spectrum(xdata[i][j]))
+
+xshape = xprocessed.shape
+numberOfItems = xshape[0]
+numberOfElectrodes = xshape[1]
+numberOfReadings = xshape[2]
+
+# plt.plot((xprocessed[6][2]), color='green', linewidth=1)
+# plt.show()
+# exit(0)
+
+ydata = ydata.transpose(1, 0)
+ydata = list(ydata[0])
+
+# reducce y data from 1, 2 to 0, 1 respectively
+for i in range(0, numberOfItems):
+    ydata[i] = int(ydata[i] / 2)
+
+ydata = np.array(ydata)
+
+print("Shape: ")
+print(xprocessed.shape)
+print(ydata.shape)
+print(ydata)
+
+print("NRAP:")
+print(nrap)
+
+X_train = xprocessed[:(numberOfItems - 20)]
+y_train = ydata[:(numberOfItems - 20)]
+X_test = xprocessed[(numberOfItems - 20):]
+y_test = ydata[(numberOfItems - 20):]
+print("Data formation: ")
+print(X_train.shape)
+print(X_test.shape)
+print(y_train.shape)
+print(y_test.shape)
+
+X_train = X_train.reshape(X_train.shape[0], numberOfReadings, numberOfReadings, 1)
+X_test = X_test.reshape(X_test.shape[0], numberOfReadings, numberOfReadings, 1)
+X_train = X_train.astype('float32')
+X_test = X_test.astype('float32')
+Y_train = np_utils.to_categorical(y_train)
+Y_test = np_utils.to_categorical(y_test)
+
+print("Data formation: ")
+print(X_train.shape)
+print(X_test.shape)
+print(Y_train.shape)
+print(Y_test.shape)
+
+# model = Sequential()
+# model.add(Convolution2D(32, 3, 3, activation='relu', input_shape=(numberOfReadings,numberOfReadings,1)))
+# model.add(Convolution2D(40, 1, activation='relu'))
+# model.add(Convolution2D(20, 1, activation='relu'))
+# model.add(Convolution2D(10, 1, activation='relu'))
+# model.add(Convolution2D(2, (numberOfReadings-2)))
+# model.add(Flatten())
+# model.add(Activation('softmax'))
+#
+# model.summary()
+#
+# model.compile(loss='categorical_crossentropy',
+#              optimizer='adam',
+#              metrics=['accuracy'])
+#
+# model.fit(X_train, Y_train, batch_size=60, epochs=4, verbose=1)
+#
+# score = model.evaluate(X_test, Y_test, verbose=0)
+#
+# print("Score:")
+# print(score)
+model = load_model('../outputmodel1/Dataset2-2.h5')
+y_pred = model.predict(X_test.reshape(X_test.shape[0], numberOfReadings, numberOfReadings, 1))
+
+print("y predict:")
+print(y_pred[:10])
+print("y test:")
+print(y_test[:10])
+
+# model.save("../outputmodel1/Dataset2-2.h5")
diff --git a/venv/src/SyntheticData.py b/venv/src/SyntheticData.py
index 028b76a..749c867 100644
--- a/venv/src/SyntheticData.py
+++ b/venv/src/SyntheticData.py
@@ -93,7 +93,8 @@ def spectrum (vector):
 
 model = Sequential()
 model.add(Convolution2D(32, 3, 3, activation='relu', input_shape=(576,576,1)))
-model.add(Convolution2D(10, 1, activation='relu'))
+model.add(Convolution2D(40, 1, activation='relu'))
+model.add(Convolution2D(20, 1, activation='relu'))
 model.add(Convolution2D(10, 1, activation='relu'))
 # model.add(Dense(500, activation='relu'))
 model.add(Convolution2D(2, 574))
@@ -111,7 +112,7 @@ def spectrum (vector):
 print("Y_train:")
 print(Y_train.shape)
 
-model.fit(X_train, Y_train, batch_size=40, epochs=2, verbose=1)
+model.fit(X_train, Y_train, batch_size=40, epochs=3, verbose=1)
 
 score = model.evaluate(X_test, Y_test, verbose=0)
 model.save('my_model.h5')
diff --git a/venv/src/my_model.h5 b/venv/src/my_model.h5
index 1a32cb4..78da49c 100644
Binary files a/venv/src/my_model.h5 and b/venv/src/my_model.h5 differ