ā† Back to WIA-AI-012

šŸŒ Federated Learning Simulator

Interactive demonstrations of federated learning concepts

šŸ“Š Federated Averaging
šŸ‘„ Client Selection
šŸ”’ Privacy Budget
šŸ“” Communication
šŸ”„ Aggregation

Federated Averaging Demonstration

Simulate the FedAvg algorithm where multiple clients train locally and updates are aggregated on the server.

Clients: 10
Rate: 50%
Epochs: 5
LR: 0.01
Current Round
0
Global Accuracy
0.0%
Participating Clients
0
Avg. Local Loss
0.00

Training Progress

Client Selection Strategies

Compare different client selection strategies and their impact on training.

Selection Criteria

Select a strategy to see details.

Expected Impact

Run selection to see impact analysis.

Selected Clients

Differential Privacy Budget Calculator

Calculate and visualize privacy budget (Īµ, Ī“) for differential privacy guarantees.

Īµ = 1.0
Ī“ = 0.00001
0
Privacy Score
Privacy Level
-
Total Budget
0.0
Budget per Query
0.000
Noise Scale
0.00

Privacy Interpretation

Adjust parameters to see privacy analysis.

Communication Efficiency Analyzer

Analyze bandwidth usage and optimization techniques for federated learning.

Total Bandwidth
0GB
Per Round
0MB
Savings
0%
Time Estimate
0min

Bandwidth Comparison

Compression Details

Select a compression method to see details.

Recommendations

Calculate bandwidth to see recommendations.

Model Aggregation Visualizer

Visualize different aggregation algorithms and their behavior.

Client Updates Distribution

Aggregated Value
0.00
Variance
0.00
Outliers Rejected
0
Robustness Score
0%

Algorithm Explanation

Select an algorithm to see how it works.