curiousily

var Mongoose = require('mongoose')
var Hapi = require('hapi')
var Co = require('co')

Mongoose.connect('mongodb://localhost/mytestdb')

var userSchema = Mongoose.Schema({
    name: String
})

curiousily

// What is a Tensor?

const myFirstTensor = tf.scalar(42)
console.log(myFirstTensor)
myFirstTensor.print()


const oneDimTensor = tf.tensor1d([1, 2, 3])
oneDimTensor.print()

curiousily

data = OrderedDict(
  amount_spent =  [50,  10, 20, 5,  95,  70,  100,  200, 0],
  send_discount = [0,   1,  1,  1,  0,   0,   0,    0,   1]
)

curiousily

df = pd.DataFrame.from_dict(data)

curiousily

def sigmoid(z):
  return 1 / (1 + np.exp(-z))

curiousily

def loss(h, y):
  return (-y * np.log(h) - (1 - y) * np.log(1 - h)).mean()

curiousily

X = df['amount_spent'].astype('float').values
y = df['send_discount'].astype('float').values

def predict(x, w):
  return sigmoid(x * w)

def print_result(y_hat, y):
  print(f'loss: {np.round(loss(y_hat, y), 5)} predicted: {y_hat} actual: {y}')
  
y_hat = predict(x=X[0], w=.5)

curiousily

for w in np.arange(-1, 1, 0.1):
  y_hat = predict(x=X[0], w=w)
  print(loss(y_hat, y[0]))

curiousily

def fit(X, y, n_iter=100000, lr=0.01):

  W = np.zeros(X.shape[1])

  for i in range(n_iter):
      z = np.dot(X, W)
      h = sigmoid(z)
      gradient = np.dot(X.T, (h - y)) / y.size
      W -= lr * gradient
  return W

curiousily

def predict(X, W):
  return sigmoid(np.dot(X, W))