Explainability
Definition
The degree to which a model's output can be explained in terms of its inputs and logic. Logistic regression and decision trees are inherently explainable; deep neural networks are not. Forensic applications prioritise explainability because the output must withstand expert cross-examination.
- Explainable models
- Logistic regression, decision trees
- Low-explainability models
- Deep neural networks
- Forensic priority reason
- Output must withstand expert cross-examination
- Field
- Fraud detection and data analytics
Common questions
Why would a forensic team choose a less accurate but more explainable model?+
An opinion that cannot be explained in terms a judge or jury can follow is vulnerable to exclusion or heavy discounting on cross-examination, so an interpretable model whose reasoning can be defended is often preferred over a marginally more accurate black box.
Does using an explainable model like a decision tree remove the need to validate it?+
No. Explainability describes whether the model's logic can be inspected and communicated, not whether it is accurate or reliable. A transparent model still needs testing against known data and a disclosed error rate before its output supports a conclusion.
Related terms
- Anomaly Scoring
- A numeric score assigned to each entity or transaction based on how different it is from the expected population, derived from multiple...
- Isolation Forest
- An unsupervised machine-learning model for anomaly detection. It builds random decision trees and scores each record by the average depth required to...
- Logistic Regression (Supervised Fraud Model)
- A classification model trained on historically labelled transactions (fraud vs. legitimate) to estimate the probability that a new transaction is fraudulent. Requires...
- Network Analysis (Link Analysis)
- A method that models entities (people, companies, accounts, addresses) as nodes and connections between them (shared attributes, transactions, ownership) as edges, then...
- Timeline Reconstruction
- The process of ordering digital events from multiple sources into a single chronological account. Requires normalising all timestamps to a common reference...