Relationship of attacks and machine learning model development process asset

General Image Description

The UML class diagram visualizes a threat model with 1 threat, determined from the conducted systematic literature review, which targets the machine learning model development process for the initial compromise. The compromise of the machine learning model development process is conducted through “Enabling software components”.

Machine learning model development process: a systematic continuous process of machine learning development, incorporating software components together to create a model, enable its training and conduct inference operations. The goal is to build a machine learning model, which will be able to solve the target business specific problem to achieve a target objective. The process may involve activities like data collection, data preprocessing, model design, building and training, quality assurance, deployment, monitoring, feedback, and auditability.

List of threats

  1. [MLMDP.T.1] A supply chain attack targets vulnerabilities in the machine learning (ML) supply chain to compromise the integrity, security, and trustworthiness of ML models and their deployment platforms. These attacks exploit weaknesses in third-party components, training data, pre-trained models, and deployment infrastructure.