Exercise

# Initialization in TensorFlow

A good initialization can reduce the amount of time needed to find the global minimum. In this exercise, we will initialize weights and biases for a neural network that will be used to predict credit card default decisions. To build intuition, we will use the low-level, linear algebraic approach, rather than making use of convenience functions and high-level `keras`

operations. We will also expand the set of input features from 3 to 23. Several operations have been imported from `tensorflow`

: `Variable()`

, `random()`

, and `ones()`

.

Instructions

**100 XP**

- Initialize the layer 1 weights,
`w1`

, as a`Variable()`

with shape`[23, 7]`

, drawn from a normal distribution. - Initialize the layer 1 bias using ones.
- Use a draw from the normal distribution to initialize
`w2`

as a`Variable()`

with shape`[7, 1]`

. - Define
`b2`

as a`Variable()`

and set its initial value to 0.0.