on_boarding.md

graph LR
    Data_Management["Data Management"]
    Geometric_Layers["Geometric Layers"]
    Wavelet_Layers["Wavelet Layers"]
    Core_Deep_Learning_Models["Core Deep Learning Models"]
    Training_Experimentation["Training & Experimentation"]
    Data_Management -- "provides processed input data to" --> Core_Deep_Learning_Models
    Geometric_Layers -- "are integrated into the architecture of" --> Core_Deep_Learning_Models
    Wavelet_Layers -- "are integrated into the architecture of" --> Core_Deep_Learning_Models
    Training_Experimentation -- "instantiates and trains" --> Core_Deep_Learning_Models
    Training_Experimentation -- "configures optimizers for" --> Core_Deep_Learning_Models
    click Data_Management href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/neuro-green/Data_Management.md" "Details"
    click Geometric_Layers href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/neuro-green/Geometric_Layers.md" "Details"
    click Wavelet_Layers href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/neuro-green/Wavelet_Layers.md" "Details"
    click Core_Deep_Learning_Models href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/neuro-green/Core_Deep_Learning_Models.md" "Details"
    click Training_Experimentation href "https://github.com/CodeBoarding/GeneratedOnBoardings/blob/main/neuro-green/Training_Experimentation.md" "Details"

Loading

Details

The neuro-green project is structured as a modular Deep Learning Research Library/Framework, focusing on computational neuroscience applications. Its architecture emphasizes clear separation of concerns for data handling, specialized neural network layers, core model definitions, and training orchestration.

Data Management [Expand]

This component is responsible for the entire data pipeline, including loading, preprocessing, and preparing neurophysiological data (e.g., EEG/MEG epochs) into a format suitable for deep learning models. It encapsulates dataset creation and any necessary data transformations.

Related Classes/Methods:

• green.data_utils (1:1000)

Geometric Layers [Expand]

This component implements specialized neural network layers and operations that operate on Symmetric Positive Definite (SPD) matrices. It leverages geometric deep learning principles to handle covariance matrices or other SPD-valued data within the neural network architecture.

Related Classes/Methods:

• green.spd_layers (1:1000)

Wavelet Layers [Expand]

This component provides custom neural network layers for extracting time-frequency features using wavelet transforms, specifically Gabor wavelets. These layers are crucial for time-frequency analysis of neurophysiological signals within the deep learning models.

Related Classes/Methods:

• green.wavelet_layers (1:1000)

Core Deep Learning Models [Expand]

This central component defines the main deep learning architectures (Green, GreenG2, GreenRegressorLM, GreenClassifierLM) built using PyTorch Lightning. These models integrate the specialized Geometric Layers and Wavelet Layers to perform specific tasks (e.g., regression, classification) on neurophysiological data. It also includes auxiliary functions directly supporting model operations.

Related Classes/Methods:

• green.research_code.pl_utils (1:1000)

Training & Experimentation [Expand]

This component manages the entire training and experimentation lifecycle of the deep learning models. It handles aspects such as cross-validation setups, optimizer configuration, and the overall execution flow for model training and evaluation using PyTorch Lightning.

Related Classes/Methods:

• green.research_code.crossval_utils (1:1000)
• green.research_code.pl_utils (1:1000)

FAQ