Rsych

import tensorflow as tf
from keras.layers import SimpleRNN
from keras.models import Sequential
import os
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'

model_rnn = Sequential()
model_rnn.add(SimpleRNN(units=20, activation='relu', input_shape=(10,5)))

print(model_rnn.summary())

Rsych

def AND(x1, x2):
    w1 = 0.5
    w2 = 0.5
    theta = 0.7
    if w1 * x1 + w2 * x2 > theta:
        return 1
    else:
        return 0
print(AND(0,0))
print(AND(0,1))

Rsych

def NAND(x1, x2):
    w1 = -0.5
    w2 = -0.5
    theta = -0.7
    if w1 * x1 + w2 * x2 > theta:
        return 1
    else:
        return 0

def OR(x1, x2):

Rsych

def stepfunc(x):
    return np.where(x <= 0, 0, 1) # if x <= 0 return 0, else return 1

x = np.arange(-10, 10, 0.1)
y = stepfunc(x)

plt.plot(x, y)
plt.title('Step Function')
plt.show()

Rsych

def sigmoid(x):
    return 1 / (1 + np.exp(-x))
x = np.arange(-10, 10, 0.1)
y = sigmoid(x)

plt.plot(x,y)
plt.title('Sigmoid Function')
plt.show()

Rsych

def tanh(x):
    return ((np.exp(2 * x) - 1) / (np.exp(2 * x) + 1) )
x = np.arange(-10, 10, 0.1)
y = tanh(x)

plt.plot(x,y)
plt.title('tanh Function')
plt.show()

Rsych

def relu(x):
    return np.maximum(0, x)

x = np.arange(-10, 10, 0.1)
y = relu(x)

plt.plot(x,y)
plt.title('ReLU Function')
plt.show()

Rsych

def softmax(x):
    return np.exp(x) / np.sum(np.exp(x))

print(softmax([1,1,3,6,7]))

Rsych

import matplotlib.pyplot as plt
import matplotlib.gridspec as gridspec
import tensorflow as tf
import seaborn as sns
sns.set()

x = [1, 2, 3, 4, 5]
y = [2, 3, 4, 5, 6]

w = tf.Variable(0.7) # we set weight to 0.7, smaller difference between initial value to actual weight is faster

Rsych

import tensorflow as tf
import numpy as np
import os
from tensorflow import keras

# Neural network and train variables
EPOCHS = 200
BATCH_SIZE = 128
VERBOSE = 1
NB_CLASSES = 10 # Num of outputs