diff --git a/spider_control/control.py b/spider_control/control.py
index f1b46da0cc8546aaf3099452efacbd5ee3d63fb7..e7ca5e0d48987bfa75acec329ab4ae6b2bd250cc 100644
--- a/spider_control/control.py
+++ b/spider_control/control.py
@@ -264,9 +264,11 @@ class MultiClassLogistic:
         global weight_hid
         global weight_out
         self.neuron_input = np.array(neuron_values)#assigning input values to neurons
+        self.neuron_hidden = np.array(4, 1)#an array to assign neuron values of hidden layer
+        self.neuron_out = np.array(3, 1)#an array to assign neuron values of output layer
         weights1 = tf.random_normal((neurons, 1), dtype=tf.float32, seed=1606)#weights of input layer
-        weights2 = tf.random_normal((neurons/4, 1), dtype=tf.float32, seed=1003)#weights for hidden layer, number of neurons in hidden layer is identified by the quadrant concept
-        weights3 = tf.random_normal((neurons, 1), dtype=tf.float32, seed=0207)#weights for output layer
+        weights2 = tf.random_normal((neurons/2, 1), dtype=tf.float32, seed=1003)#weights for hidden layer, number of neurons in hidden layer is identified by the quadrant concept
+        weights3 = tf.random_normal((3, 1), dtype=tf.float32, seed=0207)#weights for output layer
         sess = tf.Session()
         weight_in, weight_hid, weight_out = sess.run(weights1, weights2, weights3)
 
@@ -283,18 +285,40 @@ class MultiClassLogistic:
         n_input = sess.run(multiplication)
         return n_input
     #function that returns the softmax of the output layer or any given array of neurons
-    def out_softmax(self, neuron_out, neurons):
-        output = np.array(neurons)
+    def out_softmax(self, neuron_out):
+        output = np.array(3)
         output = tf.nn.softmax(neuron_out)
         return output
 
-    def activation(self, a, b):
-        'where a '
+    def output_func(self, ):
+
+    def activation(self, a):
+        'where a is the input of the neuron and the activation function is modelled as rectified linear output'
+        if(a>=0):
+            b = a
+        else:
+            b = 0
+        return b
+    def kt_learning(self, n_out = [], n_preferred = [], learning_rate):
+        'where n_out is the actual output and n_preffered is the prefered output of the network'
+        init = tf.global_variables_initializer()
+        with tf.Session as session:
+            session.run(init)
+        loss = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(labels=n_out, logits=n_preferred))
+        optimizer = tf.train.GradientDescentOptimizer(learning_rate).minimize(loss)
+        with tf.Session() as sess:
+            sess.run(optimizer)
+
+
+
+
+
+
 
 
 
 
-#function to normalize the input data
+    #function to normalize the input data
 def normalization(a):
     a_min = min(a)
     a_max = max(a)