Enum NetworkComplexity
Defines the complexity level of a neural network architecture.
public enum NetworkComplexityFields
Custom = 4Custom complexity defined by the user.
For Beginners: The Custom option allows you to define your own network structure rather than using the predefined complexity levels. This gives you complete control over how many layers your network has and how many neurons are in each layer. This option is useful when you have specific requirements or when you're experimenting to find the optimal network structure for your particular problem.
Deep = 2Deep network with many layers, suitable for complex tasks.
For Beginners: A Deep network has many layers (typically 6-10) with numerous neurons. It's like a large organization with multiple departments working together. These networks can learn complex patterns and relationships in data, making them suitable for challenging tasks like image recognition or natural language processing. They require more data and computing power but can achieve higher accuracy on difficult problems.
Medium = 1Medium complexity network with a moderate number of layers.
For Beginners: A Medium complexity network usually has 3-5 hidden layers with more neurons. It's like a medium-sized team with specialists. These networks can handle more complex patterns than Simple networks while still training in a reasonable time. They're a good balance between power and efficiency for many common problems.
Simple = 0Simple network with minimal layers, suitable for basic tasks.
For Beginners: A Simple network typically has just 1-2 hidden layers with few neurons. It's like a small team of people working on a problem. These networks train quickly, require less data, and work well for straightforward problems like linear relationships or basic classification. Use this when you have limited data or when your problem isn't very complex.
VeryDeep = 3Very deep network with extensive layers and connections.
For Beginners: A VeryDeep network has a large number of layers (10+ layers) with many neurons and connections. It's like a massive organization with highly specialized departments and complex communication channels. These networks can solve extremely complex problems and learn subtle patterns, but they require substantial data, computing power, and time to train. They might be overkill for simpler problems and can be prone to overfitting (memorizing data rather than learning patterns) if not properly managed.
Remarks
For Beginners: Neural networks are AI models inspired by the human brain. They consist of interconnected "neurons" organized in layers. The complexity of a neural network refers to how many layers it has and how many neurons are in each layer. More complex networks can learn more sophisticated patterns but require more data and computing power. This enum helps you choose an appropriate complexity level for your task without needing to understand all the technical details.