Data Version Control

Description

Data Version Control (DVC) is a Git-like version control system specifically designed for machine learning data, models, and experiments. It tracks changes to large data files, maintains reproducible ML pipelines, and creates a complete audit trail of data transformations, model training, and evaluation processes. DVC works alongside Git to provide end-to-end lineage tracking from raw data through preprocessing, training, and deployment, enabling teams to reproduce any model version and understand exactly how datasets evolved throughout the ML lifecycle.

Example Use Cases

Transparency

Tracking medical imaging dataset versions and model training pipelines to ensure reproducible research results, enabling hospitals to verify which specific data version and preprocessing steps were used for regulatory submissions.

Maintaining pharmaceutical drug discovery data lineage across multiple research teams, tracking compound datasets, feature extraction processes, and model versions to support FDA submissions with complete experimental provenance.

Reliability

Managing credit scoring model data pipelines with complete version control of training datasets, feature engineering steps, and model artifacts, ensuring reliable model reproduction and rollback capabilities when performance issues arise.

Limitations

Requires learning Git-like workflows and CLI commands, which may have a steep learning curve for teams unfamiliar with version control systems.
Storage costs can be substantial for large datasets with frequent changes, especially when maintaining multiple versions and branches of data.
Complex data pipelines with many interdependencies may require significant setup time and careful configuration to track properly.
Performance can degrade with very large files or datasets due to checksumming and synchronisation overhead during operations.
Team coordination becomes essential as improper branch management or merge conflicts can disrupt collaborative workflows.

Resources

DVC Documentation

Documentation

Comprehensive official documentation covering DVC installation, data versioning, pipeline creation, and collaborative workflows with tutorials and best practices

iterative/dvc

Software Package

Official DVC open-source repository containing the complete data version control system for machine learning with Git integration

DVC Tutorial - Data Version Control for Machine Learning

Tutorial

Step-by-step getting started guide demonstrating DVC basics including data tracking, pipeline creation, and experiment management

Related Techniques

Name	Description	Assurance Goals
Jackknife Resampling	Jackknife resampling (also called leave-one-out resampling) assesses model stability and uncertainty by systematically removing one data point at a time and retraining the model on the remaining data. Unlike bootstrapping which samples with replacement, jackknife creates n different models by excluding each of the n data points once. This systematic approach reveals how individual points influence results, provides robust estimates of prediction variance, and identifies unusually influential observations that may be outliers or leverage points affecting model reliability.	Reliability Transparency Fairness
Model Cards	Model cards are standardised documentation frameworks that systematically document machine learning models through structured templates. The templates cover intended use cases, performance metrics across different demographic groups and operating conditions, training data characteristics, evaluation procedures, limitations, and ethical considerations. They serve as comprehensive technical specifications that enable informed model selection, prevent inappropriate deployment, support regulatory compliance, and facilitate fair assessment by providing transparent reporting of model capabilities and constraints across diverse populations and scenarios.	Transparency Fairness Safety
Generalized Additive Models	An intrinsically interpretable modelling technique that extends linear models by allowing flexible, nonlinear relationships between individual features and the target whilst maintaining the additive structure that preserves transparency. Each feature's effect is modelled separately as a smooth function, visualised as a curve showing how the feature influences predictions across its range. GAMs achieve this through spline functions or other smoothing techniques that capture complex patterns in individual variables without interactions, making them particularly valuable for domains requiring both predictive accuracy and model interpretability.	Transparency Explainability Reliability
Average Odds Difference	Average Odds Difference measures fairness by calculating the average difference in both false positive rates and true positive rates between different demographic groups. This metric captures how consistently a model performs across groups for both positive and negative predictions. A value of 0 indicates perfect fairness under the equalized odds criterion, while larger absolute values indicate greater disparities in model performance between groups.	Fairness Reliability
Layer-wise Relevance Propagation	Layer-wise Relevance Propagation (LRP) explains neural network predictions by working backwards through the network to show how much each input feature contributed to the final decision. It follows a simple conservation rule: the total contribution scores always add up to the original prediction. Starting from the output, LRP distributes 'relevance' backwards through each layer using different rules depending on the layer type. This creates a detailed breakdown showing which input features helped or hindered the prediction, making it easier to understand why the network made a particular decision.	Explainability Transparency
Homomorphic Encryption	Homomorphic encryption allows computation on encrypted data without decrypting it first, producing encrypted results that, when decrypted, match the results of performing the same operations on the plaintext. This enables secure outsourced computation where sensitive data remains encrypted throughout processing. By allowing ML operations on encrypted data, it provides strong privacy guarantees for applications involving highly sensitive information.	Privacy Safety Transparency Security