Z by HP Unlocked Challenge 4 - Image Classification.
The challenge is to build a machine learning model to classify images of "La Eterna". This can be done in a variety of ways. For this tutorial we will be focusing on building an image classification using artificial neural nets (ANN).
Unlocked is an action-packed interactive film made bt Z by HP for data scientists. Sharpen your skills and solve the data driven mystery here: https://www.hp.com/us-en/workstations/industries/data-science/unlocked-challenge.html
You'll want to click "Next Challenge" until you reach Challenge 4.
The data is split into a training and a submission set. The repo includes two labeled folders in the Test folder. The folder labeled "la_eterna" includes the pictures of la eterna that Eva captured. The other folder labeled "other_flowers" includes pictures of other flowers that are not la eterna. We will use this data to build our classifier. Each of the images has been formatted to the dimensions (224,224, 3) for the analysis.
We have provided some starter code and a full tutorial video on how to approach this problem. All the models that we built can be tweaked and improved upon.
- Introduction to Neural Networks in Python | Tensorflow/Keras
- Real-World Python Neural Nets Tutorial (Image Classification w/ CNN)
- Ken's Tutorial for this Challenge
- Step by Step Guide to Image Classification with New Datasets
- Image Classification from Scratch
- Building powerful image classification models using very little data
- Keras Loss Functions
- Load and Preprocess Images
- Convolutional Neural Nets
- Simple Guide to Hyperparameter Tuning in Neural Networks
- A Comprehensive Guide to Convolutional Neural Nets (With Pooling)