# niyaz05/early-sepsis-detection-using-explainable-ai architecture

> Architecture diagram of the GitHub repository niyaz05/early-sepsis-detection-using-explainable-ai, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-11-20). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/niyaz05/early-sepsis-detection-using-explainable-ai
- Repository: https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai

## Overview

This repository is a reproducible ICU sepsis early-warning research system, not a deployed clinical service. Its user workflow runs a documented sequence of notebooks: explore and prepare patient-hour data, train and evaluate four model families, then compare performance and inspect clinical lead time, explainability, uncertainty, and fairness. Shared utilities support schema, metrics, and sequence handling; saved models and result tables capture outputs. The graph follows README-described notebook stages and sampled generator/helper references; the raw data directory is documented but absent from the supplied tree.

Main entry points: [`notebooks/02_preprocessing_and_splits.ipynb`](https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/blob/main/notebooks/02_preprocessing_and_splits.ipynb) (Patient splits and features), [`models`](https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/tree/main/models) (Saved models), [`results`](https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/tree/main/results) (Predictions and metrics), [`notebooks/03_model_xgboost.ipynb`](https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/blob/main/notebooks/03_model_xgboost.ipynb) (XGBoost training)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_data["Data preparation"]
  node_eda["Clinical EDA"]
  node_prep["Patient splits and features"]
end

subgraph group_training["Model training"]
  node_xgb["XGBoost training"]
  node_lstm["LSTM training"]
  node_tcn["TCN training<br/>[05_model_tcn.ipynb]"]
  node_transformer["Transformer training"]
  node_torch_arch["Sequence architectures<br/>[torch_models.py]"]
  node_model_files["Saved models"]
end

subgraph group_evaluation["Evaluation and analysis"]
  node_compare["Model comparison"]
  node_clinical["Clinical lead time"]
  node_xai_fairness["XAI and fairness"]
  node_result_files["Predictions and metrics"]
end

subgraph group_shared["Shared runtime"]
  node_utils["Shared data and metrics helpers<br/>[utils.py]"]
end

node_analyst(("Researcher"))
node_dataset["ICU dataset"]

node_analyst -->|"runs workflow"| node_eda
node_dataset -->|"profiles"| node_eda
node_eda -->|"provides schema"| node_prep
node_prep -->|"provides tabular data"| node_xgb
node_prep -->|"provides sequences"| node_lstm
node_prep -->|"provides sequences"| node_tcn
node_prep -->|"provides sequences"| node_transformer
node_lstm -->|"uses architecture"| node_torch_arch
node_tcn -->|"uses architecture"| node_torch_arch
node_transformer -->|"uses architecture"| node_torch_arch
node_xgb -->|"saves model"| node_model_files
node_lstm -->|"saves model"| node_model_files
node_tcn -->|"saves model"| node_model_files
node_transformer -->|"saves model"| node_model_files
node_xgb -->|"writes outputs"| node_result_files
node_lstm -->|"writes outputs"| node_result_files
node_tcn -->|"writes outputs"| node_result_files
node_transformer -->|"writes outputs"| node_result_files
node_model_files -->|"compares saved predictions"| node_compare
node_result_files -->|"loads predictions"| node_compare
node_compare -->|"evaluates lead time"| node_clinical
node_model_files -->|"loads deep checkpoints"| node_xai_fairness
node_result_files -->|"loads metrics"| node_xai_fairness
node_eda -->|"uses helpers"| node_utils
node_prep -->|"uses helpers"| node_utils
node_xgb -->|"uses helpers"| node_utils
node_xai_fairness -->|"uses sequence helpers"| node_utils

click node_eda "https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/blob/main/notebooks/01_eda_overview.ipynb"
click node_prep "https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/blob/main/notebooks/02_preprocessing_and_splits.ipynb"
click node_xgb "https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/blob/main/notebooks/03_model_xgboost.ipynb"
click node_lstm "https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/blob/main/notebooks/04_model_lstm.ipynb"
click node_tcn "https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/blob/main/notebooks/05_model_tcn.ipynb"
click node_transformer "https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/blob/main/notebooks/06_model_transformer.ipynb"
click node_torch_arch "https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/blob/main/src/torch_models.py"
click node_model_files "https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/tree/main/models"
click node_compare "https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/blob/main/notebooks/07_model_comparison.ipynb"
click node_clinical "https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/blob/main/notebooks/07_model_comparison.ipynb"
click node_xai_fairness "https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/blob/main/notebooks/08_innovation_xai_uncertainty_fairness.ipynb"
click node_result_files "https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/tree/main/results"
click node_utils "https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/blob/main/src/utils.py"

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_eda,node_prep toneBlue
class node_xgb,node_lstm,node_tcn,node_transformer,node_torch_arch,node_model_files toneAmber
class node_compare,node_clinical,node_xai_fairness,node_result_files toneMint
class node_utils toneRose
class node_analyst,node_dataset toneIndigo
```

## Components

### Data preparation

- [Clinical EDA](https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/blob/main/notebooks/01_eda_overview.ipynb): `notebooks/01_eda_overview.ipynb`
- [Patient splits and features](https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/blob/main/notebooks/02_preprocessing_and_splits.ipynb): `notebooks/02_preprocessing_and_splits.ipynb`

### Model training

- [XGBoost training](https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/blob/main/notebooks/03_model_xgboost.ipynb): `notebooks/03_model_xgboost.ipynb`
- [LSTM training](https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/blob/main/notebooks/04_model_lstm.ipynb): `notebooks/04_model_lstm.ipynb`
- [TCN training](https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/blob/main/notebooks/05_model_tcn.ipynb): `notebooks/05_model_tcn.ipynb`
- [Transformer training](https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/blob/main/notebooks/06_model_transformer.ipynb): `notebooks/06_model_transformer.ipynb`
- [Sequence architectures](https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/blob/main/src/torch_models.py): `src/torch_models.py`
- [Saved models](https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/tree/main/models): `models`

### Evaluation and analysis

- [Model comparison](https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/blob/main/notebooks/07_model_comparison.ipynb): `notebooks/07_model_comparison.ipynb`
- [Clinical lead time](https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/blob/main/notebooks/07_model_comparison.ipynb): `notebooks/07_model_comparison.ipynb`
- [XAI and fairness](https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/blob/main/notebooks/08_innovation_xai_uncertainty_fairness.ipynb): `notebooks/08_innovation_xai_uncertainty_fairness.ipynb`
- [Predictions and metrics](https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/tree/main/results): `results`

### Shared runtime

- [Shared data and metrics helpers](https://github.com/niyaz05/early-sepsis-detection-using-explainable-ai/blob/main/src/utils.py): `src/utils.py`

### Other components

- **Researcher**
- **ICU dataset**

## Connections

- Researcher → Clinical EDA: runs workflow (evidence: `README.md`)
- ICU dataset → Clinical EDA: profiles (evidence: `README.md`)
- Clinical EDA → Patient splits and features: provides schema (evidence: `README.md`)
- Patient splits and features → XGBoost training: provides tabular data (evidence: `README.md`)
- Patient splits and features → LSTM training: provides sequences (evidence: `README.md`)
- Patient splits and features → TCN training: provides sequences (evidence: `README.md`)
- Patient splits and features → Transformer training: provides sequences (evidence: `README.md`)
- LSTM training → Sequence architectures: uses architecture
- TCN training → Sequence architectures: uses architecture
- Transformer training → Sequence architectures: uses architecture
- XGBoost training → Saved models: saves model (evidence: `README.md`)
- LSTM training → Saved models: saves model
- TCN training → Saved models: saves model
- Transformer training → Saved models: saves model
- XGBoost training → Predictions and metrics: writes outputs (evidence: `README.md`)
- LSTM training → Predictions and metrics: writes outputs
- TCN training → Predictions and metrics: writes outputs
- Transformer training → Predictions and metrics: writes outputs
- Saved models → Model comparison: compares saved predictions (evidence: `README.md`)
- Predictions and metrics → Model comparison: loads predictions
- Model comparison → Clinical lead time: evaluates lead time
- Saved models → XAI and fairness: loads deep checkpoints
- Predictions and metrics → XAI and fairness: loads metrics
- Clinical EDA → Shared data and metrics helpers: uses helpers
- Patient splits and features → Shared data and metrics helpers: uses helpers
- XGBoost training → Shared data and metrics helpers: uses helpers
- XAI and fairness → Shared data and metrics helpers: uses sequence helpers
