Welcome to the AI-SDC documentation!

API Reference

This is the class and function reference for safemodel.

Contents:

• Introduction
  • What is safemodel?
• Preprocessing
  • DataNotAvailable
  • UnknownDataset
  • get_data_sklearn()
• Attacks
  • auc_p_val()
  • get_metrics()
  • get_probabilities()
  • min_max_disc()
  • WorstCaseAttack
  • main()
  • NumpyArrayEncoder
  • add_output_to_pdf()
  • create_json_report()
  • create_lr_report()
  • create_mia_report()
  • line()
  • subtitle()
  • title()
  • DummyClassifier
  • LIRAAttack
  • main()
• Base Class
  • SafeModel
  • check_equal()
  • check_max()
  • check_min()
  • check_type()
• SafeDecisionTree
  • SafeDecisionTreeClassifier
  • decision_tree_internal_trees_are_equal()
  • decision_trees_are_equal()
  • get_tree_k_anonymity()
• SafeRandomForest
  • SafeRandomForestClassifier
• SafeKerasModel
  • SafeKerasModel
  • check_DP_used()
  • check_checkpoint_equality()
  • check_optimizer_allowed()
  • check_optimizer_is_DP()
  • load_safe_keras_model()
  • same_configs()
  • same_weights()
• SafeSVC
  • SafeSVC
• Examples
  • SafeDecisionTree Full Length Tutorial
  • SafeRandomForest Quick Start Notebook
  • SafeSVC Quick Start Notebook
  • SafeKeras Quick Start Notebook
• Extending SafeModel
  • Modular Design
  • Copy The Template
  • Define the Safer Class
  • Update rules.json file
  • Update the __init__ method with paramnames, ignore_items, and examine_separately items
  • Add checks for any unusual data structures
  • Override the fit() function
  • Update Sphinx documentation
  • Write pytests to confirm core functionality
  • Include any optional helper functions
• JSON Output for MIA attacks
  • General Structure
  • Worst-Case Attack
  • LIRA Attack
• Running MIA Attacks from Config File

Indices and tables

• Index

• Module Index

• Search Page

Funding Statement

This work was funded by UK Research and Innovation under Grant Numbers MC_PC_21033 and MC_PC_23006 as part of Phase 1 of the DARE UK (Data and Analytics Research Environments UK) programme (https://dareuk.org.uk/), delivered in partnership with Health Data Research UK (HDR UK) and Administrative Data Research UK (ADR UK).. The specific projects were Semi-Automatic checking of Research Outputs (SACRO -MC_PC_23006) and Guidelines and Resources for AI Model Access from TrusTEd Research environments (GRAIMATTER - MC_PC_21033).­ This project has also been supported by MRC and EPSRC [grant number MR/S010351/1]: PICTURES.