A feed-forward neural network using Keras

Keras is a DL library, originally built on Python, that runs over TensorFlow or Theano. It was developed to make DL implementations faster:

1. We call install keras using the following command in your operation system's Command Prompt:

pip install keras

2. We start by importing the numpy and pandas library for data manipulation. Also, we set a seed that allows us to reproduce the script's results:

import numpy as np
import pandas as pd
numpy.random.seed(8)

3. Next, the sequential model and dense layers are imported from keras.models and keras.layers respectively. Keras models are defined as a sequence of layers. The sequential construct allows the user to configure and add layers. The dense layer allows a user to build a fully connected network:

from keras.models import Sequential
from keras.layers import Dense

4. The HR attrition dataset is then loaded, which has 14,999 rows and 10 columns. The salary and sales attributes are one-hot encoded to be used by Keras for...