From 2540d1a0866d3b73473585a3efd415c7f3104195 Mon Sep 17 00:00:00 2001
From: Karthik Udyawar <54676225+KarthikUdyawar@users.noreply.github.com>
Date: Sun, 1 Sep 2019 19:54:11 +0530
Subject: [PATCH] New Updates and fixes

+ New Training inputs and outputs.
+ Constants like iterations and learning rate.
+ Improve sigmoid function.
+ And some miner changes.
---
 main-1.1.py | 62 +++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 62 insertions(+)
 create mode 100644 main-1.1.py

diff --git a/main-1.1.py b/main-1.1.py
new file mode 100644
index 0000000..a1012c1
--- /dev/null
+++ b/main-1.1.py
@@ -0,0 +1,62 @@
+import numpy as np
+
+class NeuralNetwork():
+    def __init__(self):
+        np.random.seed(1)
+
+        self.learning_rate = 10
+
+        self.synaptic_weights = 2 * np.random.standard_normal((2, 1)) - 1
+
+    def sigmoid(self, x, deriv = False):
+        if deriv == True:
+            return (1 / (1 + np.exp(-x))) * (1 - (1 / (1 + np.exp(-x))))
+        return 1 / (1 + np.exp(-x))
+
+    def train(self, training_set_inputs, training_set_outputs, number_of_training_iterations):
+        for iteration in range(number_of_training_iterations):
+            output = self.think(training_set_inputs)
+
+            error = training_set_outputs - output
+
+            adjustment = np.dot(training_set_inputs.T, error * self.sigmoid(output,deriv=True))
+
+            self.synaptic_weights += adjustment * self.learning_rate
+
+    def think(self, inputs):
+        return self.sigmoid(np.dot(inputs, self.synaptic_weights))
+
+if __name__ == "__main__":
+
+    iterations = 10000
+
+    neural_network = NeuralNetwork()
+
+    print ("Random starting synaptic weights: ")
+    print (neural_network.synaptic_weights)
+
+    training_set_inputs = np.array(([[3,1.5],
+                                     [2,1],
+                                     [4,1.5],
+                                     [3,1],
+                                     [4,0.5],
+                                     [2,0.5],
+                                     [5.5,1],
+                                     [1,1]]))
+
+    training_set_outputs = np.array([[1],
+                                     [0],
+                                     [1],
+                                     [0],
+                                     [1],
+                                     [0],
+                                     [1],
+                                     [0]])
+
+    neural_network.train(training_set_inputs, training_set_outputs, iterations)
+
+    print ("New synaptic weights after training: ")
+    print (neural_network.synaptic_weights)
+
+    print ("Considering new situation [4.5, 1] -> ?: ")
+    print (neural_network.think(np.array([[4.5, 1]])))