How ahmad-nadeem-official/ml-studio works
ML Studio is a Streamlit no-code dashboard for loading tabular datasets, exploring them, and training classification or regression models. Its primary workflow runs through a single application file: upload and inspect data, choose a model, tune and adjust parameters, train, review metrics, and download the result. The README also describes EDA visualizations, SVM scaling, and preprocessing guidance. The sampled code shows these responsibilities largely embedded in the app; separate files under main_fnx contain algorithm-specific code, but the excerpts do not establish that the app imports or dispatches to them.
Main entry points: main.py Hyperparameter tuning
Components 21
Dashboard and data exploration
- Streamlit app
main/main.py - Uploaded dataset
main/main.py - CSV / Excel loading
main/main.py - Data review
main/main.py - EDA visualizations
main/main.py
Modeling workflow
- Task and model selection
main/main.py - Split and preprocessing
main/main.py - Classification models
main/main.py - Regression models
main/main.py - Hyperparameter tuning
main/main.py - Final model training
main/main.py - Random Forest code
main_fnx/RF.py - SVM code
main_fnx/SVM.PY - KNN code
main_fnx/KNN.py - Logistic regression code
main_fnx/logistic.py - Linear regression code
main_fnx/linear.py
Evaluation and export
- Model evaluation
main/main.py - Model download
main/main.py
Other components
- Dashboard user
- Optuna
- Colab cleaning guide
Connections 19
- Dashboard user → Streamlit app: uses
- Dashboard user → Uploaded dataset: uploads
- Streamlit app → CSV / Excel loading: loads
- Uploaded dataset → CSV / Excel loading: provides file
- CSV / Excel loading → Data review: provides data
- Data review → EDA visualizations: explores
- Dashboard user → Task and model selection: selects task
- Task and model selection → Split and preprocessing: configures data
- Split and preprocessing → Classification models: supplies data
- Split and preprocessing → Regression models: supplies data
- Classification models → Hyperparameter tuning: offers parameters
- Regression models → Hyperparameter tuning: offers parameters
- Hyperparameter tuning → Optuna: optimizes with
- Hyperparameter tuning → Final model training: provides parameters
- Dashboard user → Final model training: starts training
- Final model training → Model evaluation: produces predictions
- Model evaluation → Model download: presents model
- Dashboard user → Model download: downloads
- Colab cleaning guide → Dashboard user: guides cleaning