Cats vs. Dogs Image Classifier with a CNN (TensorFlow + Keras) · Matt Alexius

A convolutional neural network that tells cats from dogs in photos. Built with Python, TensorFlow, and Keras, with training charts in Matplotlib.

A machine learning project that sorts photos of cats and dogs with a convolutional neural network. I built it in Python with TensorFlow and Keras and trained the model to pick up the features that tell the two apart. Plotting training accuracy, loss, and predictions in Matplotlib helped me see how a deep learning model handles image data.

What I learned

CNNs look intimidating until you build one yourself. Watching the loss go down while accuracy goes up is still one of my favorite things in this field. I also learned that data quality matters more than model depth: a smaller model on a clean dataset beats a deeper one on a noisy dataset every time.