In this chapter, we have seen how to develop a real-life application using CNNs on the DL4J framework. We have seen how to solve a multi-label classification problem through nine CNNs and a series of complex feature engineering and image manipulation operations. Albeit, we couldn't achieve higher accuracy, but readers are encouraged to tune hyperparameters in the code and try the same approach with the same dataset.
Also, training the CNNs with all the images is recommended so that networks can get enough data to learn the features from Yelp images. One more suggestion is improving the feature extraction process so that the CNNs can have more quality features.
In the next chapter, we will see how to implement and deploy a hands-on deep learning project that classifies review texts as either positive or negative based on the words they contain. A large-scale movie...
