Functional API

As explained in Deep Learning with R, Third Edition

The Keras Functional API is an interface for constructing models by connecting layer operations as a directed computation graph.

The Keras Functional API constructs models by connecting layer operations as a directed graph. It extends beyond Sequential models, which are limited to simple one-input, one-output stacks. section 3.6.3 section 7.2.2

A two-layer Functional model

inputs <- keras_input(shape = c(3), name = "my_input") features <- inputs |> layer_dense(64, activation = "relu") outputs <- features |> layer_dense(10, activation = "softmax") model <- keras_model(inputs = inputs, outputs = outputs, name = "my_functional_model") keras_input() creates a symbolic tensor containing shape and dtype specifications rather than data. Calling layers on it produces new symbolic tensors; keras_model() then binds the declared inputs and outputs. section 7.2.2

Use the API when the model requires graph structure rather than a flat layer sequence:

• Combine Functional models with Sequential components, subclassed layers, or subclassed models. section 7.2.4

Figure 7.1 Progressive disclosure of complexity for model building

Compare Sequential, Functional API, and model subclassing as progressively more flexible model-building workflows. section 7.2

How Deep Learning with R, Third Edition covers it

How deep it goes

Explains symbolic tensors, layer connections, model construction, inspection, plotting, and intermediate reuse, but defers full training-loop implementation.

Examples

Demonstrates a two-Dense model, a ticket-priority classifier, model plots, intermediate-output reuse, and mixed Functional-subclassed models.

What it assumes

Assumes familiarity with Sequential models, Dense layers, and compile(), fit(), evaluate(), and predict(); teaches symbolic tensors explicitly.

What you can do after

Build it from scratch by declaring symbolic inputs, connecting layers, binding outputs, training with named inputs, and inspecting or reusing graph components.

How it connects

Connects Sequential models, model subclassing, symbolic tensors, multi-task learning, feature extraction, and model visualization.

liveBookmanning.com

Read Deep Learning with R, Third Edition on liveBook Get the book at manning.com

Also explained in

• 

  liveBookmanning.com
  Deep Learning with Python, Third EditionThe Keras model-building interface that expresses a model as a directed acyclic graph of layers with explicit connectivity.
• 

  liveBookmanning.com
  Evolutionary Deep Learning: Genetic algorithms and neural networksThe Keras Functional API is an interface for defining neural-network models as directed graphs of layer operations and tensors.

Compare how 3 books explain Functional API →

Related concepts

• Sequential modelalternative to · A sequential model is a neural-network model comprising a linear, ordered stack of layers through which data flows from input to output.
• deep learningused by · Deep learning is a machine-learning approach that uses neural networks with multiple computational layers to learn representations from data.
• model subclassingalternative to · Model subclassing is a technique for defining a neural-network model by creating a class that extends a model base class and specifies its computation.
• directed acyclic graphuses · A directed acyclic graph is a graph with directed edges and no directed cycles.
• graphuses · A graph is a mathematical structure consisting of vertices connected by edges that represent relationships or connections.

This explainer was generated by Manning's knowledge graph from the text of Deep Learning with R, Third Edition. The section links point to the parts of the book it draws on.