# oeallen-umass/ece535-federated-learning-bias-analysis architecture

> Architecture diagram of the GitHub repository oeallen-umass/ece535-federated-learning-bias-analysis, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-11-17). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/oeallen-umass/ece535-federated-learning-bias-analysis
- Repository: https://github.com/oeallen-umass/ece535-federated-learning-bias-analysis

## Overview

This repository is a small federated-learning experiment centered on FashionMNIST and FedAvg. The documented workflow partitions training data among simulated clients, trains local model copies, aggregates their weights, then evaluates the global model and compares client-group bias and accuracy variance. The README also describes TERM, AFL, and visualizations, but the supplied tree contains no dedicated implementation files for those capabilities; those branches are documented scope, not confirmed runtime modules.

Main entry points: [`src/fedavg.py`](https://github.com/oeallen-umass/ece535-federated-learning-bias-analysis/blob/main/src/fedavg.py) (FedAvg Training), [`src/evaluate.py`](https://github.com/oeallen-umass/ece535-federated-learning-bias-analysis/blob/main/src/evaluate.py) (Accuracy Evaluation), [`src/data.py`](https://github.com/oeallen-umass/ece535-federated-learning-bias-analysis/blob/main/src/data.py) (FashionMNIST Data), [`src/client.py`](https://github.com/oeallen-umass/ece535-federated-learning-bias-analysis/blob/main/src/client.py) (Simulated Clients)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_data["Data and Clients"]
  node_fashion_data[("FashionMNIST Data<br/>[data.py]")]
  node_partitioning["Data Partitioning<br/>[partition.py]"]
  node_clients["Simulated Clients<br/>[client.py]"]
end

subgraph group_learning["Federated Learning"]
  node_fedavg_runner["FedAvg Training<br/>[fedavg.py]"]
  node_aggregation["Weighted Averaging<br/>[fedavg.py]"]
  node_term_afl["TERM and AFL"]
end

subgraph group_analysis["Evaluation and Analysis"]
  node_evaluation["Accuracy Evaluation<br/>[evaluate.py]"]
  node_bias_analysis["Bias and Variance"]
  node_visualization["Result Visualizations"]
end

subgraph group_model["Model"]
  node_cnn["FashionCNN<br/>[model.py]"]
end

node_researcher(("Experimenter"))

node_researcher -->|"starts training"| node_fedavg_runner
node_fedavg_runner -->|"loads datasets"| node_fashion_data
node_fedavg_runner -->|"partitions data"| node_partitioning
node_partitioning -->|"reads labels"| node_fashion_data
node_fedavg_runner -->|"creates model"| node_cnn
node_fedavg_runner -->|"trains clients"| node_clients
node_clients -->|"trains local copy"| node_cnn
node_clients -->|"uses assigned data"| node_fashion_data
node_fedavg_runner -->|"aggregates weights"| node_aggregation
node_aggregation -->|"returns averaged weights"| node_fedavg_runner
node_fedavg_runner -->|"evaluates global model"| node_evaluation
node_evaluation -->|"uses test data"| node_fashion_data
node_evaluation -->|"evaluates predictions"| node_cnn
node_evaluation -.->|"supports comparison"| node_bias_analysis
node_bias_analysis -.->|"visualizes results"| node_visualization
node_fedavg_runner -.->|"compares methods"| node_term_afl

click node_fashion_data "https://github.com/oeallen-umass/ece535-federated-learning-bias-analysis/blob/main/src/data.py"
click node_partitioning "https://github.com/oeallen-umass/ece535-federated-learning-bias-analysis/blob/main/src/partition.py"
click node_clients "https://github.com/oeallen-umass/ece535-federated-learning-bias-analysis/blob/main/src/client.py"
click node_cnn "https://github.com/oeallen-umass/ece535-federated-learning-bias-analysis/blob/main/src/model.py"
click node_fedavg_runner "https://github.com/oeallen-umass/ece535-federated-learning-bias-analysis/blob/main/src/fedavg.py"
click node_aggregation "https://github.com/oeallen-umass/ece535-federated-learning-bias-analysis/blob/main/src/fedavg.py"
click node_term_afl "https://github.com/oeallen-umass/ece535-federated-learning-bias-analysis/tree/main/src"
click node_evaluation "https://github.com/oeallen-umass/ece535-federated-learning-bias-analysis/blob/main/src/evaluate.py"
click node_bias_analysis "https://github.com/oeallen-umass/ece535-federated-learning-bias-analysis/tree/main/src"
click node_visualization "https://github.com/oeallen-umass/ece535-federated-learning-bias-analysis/tree/main/src"

classDef toneNeutral fill:#f8fafc,stroke:#334155,stroke-width:1.5px,color:#0f172a
classDef toneBlue fill:#dbeafe,stroke:#2563eb,stroke-width:1.5px,color:#172554
classDef toneAmber fill:#fef3c7,stroke:#d97706,stroke-width:1.5px,color:#78350f
classDef toneMint fill:#dcfce7,stroke:#16a34a,stroke-width:1.5px,color:#14532d
classDef toneRose fill:#ffe4e6,stroke:#e11d48,stroke-width:1.5px,color:#881337
classDef toneIndigo fill:#e0e7ff,stroke:#4f46e5,stroke-width:1.5px,color:#312e81
classDef toneTeal fill:#ccfbf1,stroke:#0f766e,stroke-width:1.5px,color:#134e4a
class node_fashion_data,node_partitioning,node_clients toneBlue
class node_fedavg_runner,node_aggregation,node_term_afl toneAmber
class node_evaluation,node_bias_analysis,node_visualization toneMint
class node_cnn toneRose
class node_researcher toneIndigo
```

## Components

### Data and Clients

- [FashionMNIST Data](https://github.com/oeallen-umass/ece535-federated-learning-bias-analysis/blob/main/src/data.py): `src/data.py`
- [Data Partitioning](https://github.com/oeallen-umass/ece535-federated-learning-bias-analysis/blob/main/src/partition.py): `src/partition.py`
- [Simulated Clients](https://github.com/oeallen-umass/ece535-federated-learning-bias-analysis/blob/main/src/client.py): `src/client.py`

### Federated Learning

- [FedAvg Training](https://github.com/oeallen-umass/ece535-federated-learning-bias-analysis/blob/main/src/fedavg.py): `src/fedavg.py`
- [Weighted Averaging](https://github.com/oeallen-umass/ece535-federated-learning-bias-analysis/blob/main/src/fedavg.py): `src/fedavg.py`
- [TERM and AFL](https://github.com/oeallen-umass/ece535-federated-learning-bias-analysis/tree/main/src): `src`

### Evaluation and Analysis

- [Accuracy Evaluation](https://github.com/oeallen-umass/ece535-federated-learning-bias-analysis/blob/main/src/evaluate.py): `src/evaluate.py`
- [Bias and Variance](https://github.com/oeallen-umass/ece535-federated-learning-bias-analysis/tree/main/src): `src`
- [Result Visualizations](https://github.com/oeallen-umass/ece535-federated-learning-bias-analysis/tree/main/src): `src`

### Model

- [FashionCNN](https://github.com/oeallen-umass/ece535-federated-learning-bias-analysis/blob/main/src/model.py): `src/model.py`

### Other components

- **Experimenter**

## Connections

- Experimenter → FedAvg Training: starts training (evidence: `src/fedavg.py`)
- FedAvg Training → FashionMNIST Data: loads datasets (evidence: `src/fedavg.py`)
- FedAvg Training → Data Partitioning: partitions data (evidence: `src/fedavg.py`)
- Data Partitioning → FashionMNIST Data: reads labels (evidence: `src/partition.py`)
- FedAvg Training → FashionCNN: creates model (evidence: `src/fedavg.py`)
- FedAvg Training → Simulated Clients: trains clients (evidence: `src/fedavg.py`)
- Simulated Clients → FashionCNN: trains local copy (evidence: `src/client.py`)
- Simulated Clients → FashionMNIST Data: uses assigned data (evidence: `src/client.py`)
- FedAvg Training → Weighted Averaging: aggregates weights (evidence: `src/fedavg.py`)
- Weighted Averaging → FedAvg Training: returns averaged weights (evidence: `src/fedavg.py`)
- FedAvg Training → Accuracy Evaluation: evaluates global model (evidence: `src/fedavg.py`)
- Accuracy Evaluation → FashionMNIST Data: uses test data (evidence: `src/evaluate.py`)
- Accuracy Evaluation → FashionCNN: evaluates predictions (evidence: `src/evaluate.py`)
- Accuracy Evaluation → Bias and Variance: supports comparison (evidence: `README.md`)
- Bias and Variance → Result Visualizations: visualizes results (evidence: `README.md`)
- FedAvg Training → TERM and AFL: compares methods (evidence: `README.md`)
