Image classification with convolutional networks

Let's look at a more realistic case for using CNNs; we will use the Stanford Dogs versus Cats dataset. This dataset has 100+ images of dogs and cats.

Note

You can download this dataset (100 images each) from the following location: https://s3.amazonaws.com/neural-networking-book/ch04/dogs_vs_cats.tar.gz

1. Import the relevant functions and Python classes:

import matplotlib.pyplot as plt
import tensorflow as tf
import pandas as pd
import numpy as np
from sklearn.metrics import confusion_matrix
import time
from datetime import timedelta
import math
import dataset
import random

2. We will define the parameters for the convolution layers. There are three convolution layers with the following parameters:

Layer number	Layer type	Number of filters/neurons
1	Convolution	32 filters
2	Convolution	32 filters
3	Convolution	64 filters
4	Fully connected	128 neurons

The Network topolgy can be represented as shown in the following diagram:

The following code should be helpful for understanding...