FeaturedUdemy
Advanced Deep Learning with Keras
4.8(236)34k studentsINTERMEDIATEEnglish
Korshub's take
The Korshub Take
This intermediate deep learning course unpacks the robust capabilities of the Keras functional API. It moves past sequential models to tackle complex network topologies, providing a strong foundation for building sophisticated neural network architectures with unconventional data flows.
Who It's For
Data scientists and deep learning practitioners who understand basic modeling and are ready to tackle complex architectures, multiple inputs, and diverse output requirements using Keras.
Key Takeaways
- The Functional API: Transition from simple sequential models to building, fitting, and predicting with the highly flexible Keras functional API.
- Complex Architectures: Learn to design networks with multiple inputs and shared layers, utilizing categorical embeddings and merge layers.
- Multi-task Learning: Construct multiple-output networks capable of simultaneously handling multi-target regression or combining regression and classification tasks.
- Model Evaluation: Gain hands-on experience interpreting model topology using summary and plot functions, and evaluate advanced models using multiple metrics.
About this course
Keras functional API
In this course, you will learn how to solve complex problems using the Keras functional API.Beginning with an introduction, you will build simple functional networks, fit them to data, and make predictions. You will also learn how to construct models with multiple inputs and a single output and share weights between layers.
Multiple-input networks
As you progress, explore building two-input networks using categorical embeddings, shared layers, and merge layers. These are the foundational building blocks for designing neural networks with complex data flows.It extends these concepts to models with three or more inputs, helping you understand the parameters and topology of your neural networks using Keras' summary and plot functions.
Multiple-output networks
In the final interactive exercises, you'll work with multiple-output networks, which can solve regression problems with multiple targets and even handle both regression and classification tasks simultaneously.By the end of the course, you'll have practical experience with advanced deep learning techniques to advance your career as a data scientist, including evaluating your models on new data using multiple metrics.
Advertisement
