# fmata17/ml_lab architecture

> Architecture diagram of the GitHub repository fmata17/ml_lab, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-11-12). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/fmata17/ml_lab
- Repository: https://github.com/fmata17/ml_lab

## Overview

This is a machine-learning study repository, not a packaged application: notebooks explore datasets, compare modeling approaches, and investigate specialized topics, with small shared Python utilities. The README describes a repeatable exploration-to-modeling-to-evaluation workflow, while the tree contains varied notebooks rather than the uniform per-dataset layout it describes. Edges reflect that documented workflow; notebook-specific connections are omitted where source was not sampled.

Main entry points: [`README.md`](https://github.com/fmata17/ml_lab/blob/main/README.md) (Data exploration), [`notebooks/iris`](https://github.com/fmata17/ml_lab/tree/main/notebooks/iris) (Iris studies), [`notebooks/mnist`](https://github.com/fmata17/ml_lab/tree/main/notebooks/mnist) (MNIST studies), [`notebooks/titanic/data_exploration.ipynb`](https://github.com/fmata17/ml_lab/blob/main/notebooks/titanic/data_exploration.ipynb) (Titanic exploration)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_workflow["Experiment workflow"]
  node_explore["Data exploration<br/>[README.md]"]
  node_preprocess["Preprocessing and splits<br/>[README.md]"]
  node_model["Model experiments<br/>[README.md]"]
  node_evaluate["Evaluation and analysis<br/>[README.md]"]
end

subgraph group_shared["Shared utilities"]
  node_eda_helpers["EDA helpers<br/>[eda_tools.py]"]
  node_dataset_helpers["Dataset utilities<br/>[utils.py]"]
end

subgraph group_notebooks["Dataset studies"]
  node_iris["Iris studies"]
  node_mnist["MNIST studies"]
  node_titanic["Titanic exploration"]
  node_vision["Face recognition"]
end

subgraph group_special["Specialized studies"]
  node_music["Music generation"]
  node_debias["Debiasing and VAE"]
  node_finetune["LLM fine-tuning"]
  node_attention["Transformer attention"]
  node_continual["Continual learning"]
  node_cuda["CUDA learning"]
  node_brain["Neural networks"]
  node_linear["Linear algebra demo<br/>[image_detector.py]"]
end

node_learner(("Researcher"))

node_learner -->|"studies data"| node_explore
node_explore -->|"prepares data"| node_preprocess
node_preprocess -->|"feeds models"| node_model
node_model -->|"evaluates results"| node_evaluate
node_learner -.->|"explores and models"| node_iris
node_learner -.->|"explores and models"| node_mnist
node_learner -.->|"explores data"| node_titanic
node_learner -.->|"studies recognition"| node_vision
node_learner -.->|"experiments with generation"| node_music
node_learner -.->|"studies debiasing"| node_debias
node_learner -.->|"fine-tunes models"| node_finetune
node_learner -.->|"studies attention"| node_attention
node_learner -.->|"studies continual learning"| node_continual
node_learner -.->|"learns CUDA"| node_cuda
node_learner -.->|"studies neural networks"| node_brain
node_learner -.->|"runs demo"| node_linear

click node_explore "https://github.com/fmata17/ml_lab/blob/main/README.md"
click node_preprocess "https://github.com/fmata17/ml_lab/blob/main/README.md"
click node_model "https://github.com/fmata17/ml_lab/blob/main/README.md"
click node_evaluate "https://github.com/fmata17/ml_lab/blob/main/README.md"
click node_eda_helpers "https://github.com/fmata17/ml_lab/blob/main/shared/eda_tools.py"
click node_dataset_helpers "https://github.com/fmata17/ml_lab/blob/main/shared/utils.py"
click node_iris "https://github.com/fmata17/ml_lab/tree/main/notebooks/iris"
click node_mnist "https://github.com/fmata17/ml_lab/tree/main/notebooks/mnist"
click node_titanic "https://github.com/fmata17/ml_lab/blob/main/notebooks/titanic/data_exploration.ipynb"
click node_vision "https://github.com/fmata17/ml_lab/tree/main/notebooks/celebA_imageNet"
click node_music "https://github.com/fmata17/ml_lab/blob/main/mit_dl/lab1_music_generation.ipynb"
click node_debias "https://github.com/fmata17/ml_lab/tree/main/mit_dl/lab2"
click node_finetune "https://github.com/fmata17/ml_lab/blob/main/mit_dl/lab3_LLM_Finetuning_Solution.ipynb"
click node_continual "https://github.com/fmata17/ml_lab/blob/main/research/E1/CL_SGD_Split-MNIST.ipynb"
click node_cuda "https://github.com/fmata17/ml_lab/blob/main/nvidia_dl_institute/An_Even_Easier_Introduction_to_CUDA.ipynb"
click node_brain "https://github.com/fmata17/ml_lab/blob/main/nvidia_dl_institute/BuildingABrain.ipynb"
click node_linear "https://github.com/fmata17/ml_lab/blob/main/notebooks/linear_algebra/image_detector.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_explore,node_preprocess,node_model,node_evaluate toneBlue
class node_eda_helpers,node_dataset_helpers toneAmber
class node_iris,node_mnist,node_titanic,node_vision toneMint
class node_music,node_debias,node_finetune,node_attention,node_continual,node_cuda,node_brain,node_linear toneRose
class node_learner toneIndigo
```

## Components

### Experiment workflow

- [Data exploration](https://github.com/fmata17/ml_lab/blob/main/README.md): `README.md`
- [Preprocessing and splits](https://github.com/fmata17/ml_lab/blob/main/README.md): `README.md`
- [Model experiments](https://github.com/fmata17/ml_lab/blob/main/README.md): `README.md`
- [Evaluation and analysis](https://github.com/fmata17/ml_lab/blob/main/README.md): `README.md`

### Shared utilities

- [EDA helpers](https://github.com/fmata17/ml_lab/blob/main/shared/eda_tools.py): `shared/eda_tools.py`
- [Dataset utilities](https://github.com/fmata17/ml_lab/blob/main/shared/utils.py): `shared/utils.py`

### Dataset studies

- [Iris studies](https://github.com/fmata17/ml_lab/tree/main/notebooks/iris): `notebooks/iris`
- [MNIST studies](https://github.com/fmata17/ml_lab/tree/main/notebooks/mnist): `notebooks/mnist`
- [Titanic exploration](https://github.com/fmata17/ml_lab/blob/main/notebooks/titanic/data_exploration.ipynb): `notebooks/titanic/data_exploration.ipynb`
- [Face recognition](https://github.com/fmata17/ml_lab/tree/main/notebooks/celebA_imageNet): `notebooks/celebA_imageNet`

### Specialized studies

- [Music generation](https://github.com/fmata17/ml_lab/blob/main/mit_dl/lab1_music_generation.ipynb): `mit_dl/lab1_music_generation.ipynb`
- [Debiasing and VAE](https://github.com/fmata17/ml_lab/tree/main/mit_dl/lab2): `mit_dl/lab2`
- [LLM fine-tuning](https://github.com/fmata17/ml_lab/blob/main/mit_dl/lab3_LLM_Finetuning_Solution.ipynb): `mit_dl/lab3_LLM_Finetuning_Solution.ipynb`
- **Transformer attention**
- [Continual learning](https://github.com/fmata17/ml_lab/blob/main/research/E1/CL_SGD_Split-MNIST.ipynb): `research/E1/CL_SGD_Split-MNIST.ipynb`
- [CUDA learning](https://github.com/fmata17/ml_lab/blob/main/nvidia_dl_institute/An_Even_Easier_Introduction_to_CUDA.ipynb): `nvidia_dl_institute/An_Even_Easier_Introduction_to_CUDA.ipynb`
- [Neural networks](https://github.com/fmata17/ml_lab/blob/main/nvidia_dl_institute/BuildingABrain.ipynb): `nvidia_dl_institute/BuildingABrain.ipynb`
- [Linear algebra demo](https://github.com/fmata17/ml_lab/blob/main/notebooks/linear_algebra/image_detector.py): `notebooks/linear_algebra/image_detector.py`

### Other components

- **Researcher**

## Connections

- Researcher → Data exploration: studies data (evidence: `README.md`)
- Data exploration → Preprocessing and splits: prepares data (evidence: `README.md`)
- Preprocessing and splits → Model experiments: feeds models (evidence: `README.md`)
- Model experiments → Evaluation and analysis: evaluates results (evidence: `README.md`)
- Researcher → Iris studies: explores and models
- Researcher → MNIST studies: explores and models
- Researcher → Titanic exploration: explores data
- Researcher → Face recognition: studies recognition
- Researcher → Music generation: experiments with generation
- Researcher → Debiasing and VAE: studies debiasing
- Researcher → LLM fine-tuning: fine-tunes models
- Researcher → Transformer attention: studies attention
- Researcher → Continual learning: studies continual learning
- Researcher → CUDA learning: learns CUDA
- Researcher → Neural networks: studies neural networks
- Researcher → Linear algebra demo: runs demo