How francescocartelli/machine-learning-put-together works
This repository is a Python toolkit for classification, preprocessing, model evaluation, and result presentation. The main workflow is to supply feature matrices and labels, configure a Graph to run preprocessing and classifiers, evaluate scores against labels, and print or plot results. Graph scheduling is the internal orchestration layer: it dispatches fitting, training, transformation, and output collection according to node type. The README documents broader classifier and graph capabilities; some implementations were not sampled, so only documented or excerpt-supported relationships are shown.
Main entry points: graphs.py Graph scheduler · utils.py Data utilities · pca.py PCA · lda.py LDA
Components 17
Graph orchestration
- Graph scheduler
src/graphs/graphs.py - Parameter grids
src/graphs/grids.py
Preprocessing and models
- PCA
src/preprocessing/pca.py - LDA
src/preprocessing/lda.py - Standardization
src/preprocessing/standard.py - Gaussianization
src/preprocessing/gaussianization.py - Gaussian classifiers
src/classifiers/gaussian_c.py - GMM classifier
src/classifiers/gmm.py - Logistic regression
src/classifiers/logistic_reg.py - SVM and kernels
src/classifiers/svm.py - Score stacking
src/transformations/transformations.py - Data utilities
src/utils.py
Evaluation and presentation
- Prediction metrics
src/measuring_predictions/measuring_predictions.py - Metrics printer
src/plotting/printing.py - Plots
src/plotting/plotting.py
Other components
- Engineer
- Feature data
Connections 22
- Engineer → Graph scheduler: configures
README.md - Engineer → Feature data: supplies
README.md - Feature data → Graph scheduler: training inputs
src/graphs/graphs.py - Graph scheduler → PCA: dispatches transform
src/graphs/graphs.py - Graph scheduler → LDA: dispatches transform
src/graphs/graphs.py - Graph scheduler → Standardization: dispatches transform
src/graphs/graphs.py - Graph scheduler → Gaussianization: dispatches transform
src/graphs/graphs.py - Graph scheduler → Gaussian classifiers: trains and scores
README.md - Graph scheduler → GMM classifier: trains and scores
src/graphs/graphs.py - Graph scheduler → Logistic regression: trains and scores
README.md - Graph scheduler → SVM and kernels: trains and scores
README.md - Graph scheduler → Parameter grids: expands configurations
src/graphs/graphs.py - Graph scheduler → Score stacking: applies transform
src/transformations/transformations.py - Graph scheduler → Metrics printer: dispatches output
src/graphs/graphs.py - Graph scheduler → Plots: dispatches output
- Metrics printer → Prediction metrics: computes DCF
src/plotting/printing.py - Plots → Prediction metrics: computes curves
src/plotting/plotting.py - GMM classifier → Data utilities: uses density helpers
src/classifiers/gmm.py - Gaussian classifiers → Data utilities: uses statistics helpers
src/classifiers/gaussian_c.py - SVM and kernels → Data utilities: uses array helpers
src/classifiers/svm.py - PCA → Data utilities: uses covariance helper
src/preprocessing/pca.py - LDA → Data utilities: uses matrix helper
src/preprocessing/lda.py