Sensitivity analysis
Sensitivity analysis is a method for examining how changes in inputs, parameters, assumptions, or conditions affect analytical results or model behavior.
Build AI Drug Discovery Pipelines
Sensitivity analysis evaluates how changes in inputs or assumptions affect a model's outputs.
- How it is used
- Uses descriptor perturbations to interpret predictions across machine-learning models.
- How deep it goes
- Defines global and local procedures, then implements local derivatives with central finite differences.
- Examples
- Analyzes descriptors for a single RNA binder from the ROBIN data set and visualizes derivatives above 0.01.
- What you can do after
- Diagnose descriptor influence by perturbing features and plotting estimated prediction sensitivities.
Read how Build AI Drug Discovery Pipelines explains sensitivity analysis →
Applied Reinforcement Learning
An analysis of how changes in model inputs or parameters affect model outputs or objectives.
- How it is used
- Uses parameter changes to assess their effects on a business optimization model’s objective.
- How deep it goes
- Defines the purpose as checking parameter effects on the objective, without detailing procedures or methods.
- Examples
- Uses raw-material prices as external parameters in a stock-portfolio model.
- What you can do after
- Evaluate how changes in external parameter values affect the objective function.
Read how Applied Reinforcement Learning explains sensitivity analysis →
Also covered in
- Causal AI
- Statistics Every Programmer Needs
- Architecting for AutonomyIn early access (MEAP)
Related concepts
- refutationincludes · Refutation is the process of challenging a causal model or causal conclusion by comparing its implications with evidence or observed intervention outcomes.
- layer sensitivity analysisincludes · Layer sensitivity analysis measures how much a model's output quality changes when individual layers are quantized or otherwise perturbed.
- unobserved confoundingmeasures · Unobserved confounding is confounding caused by a variable that influences both treatment and outcome but is not measured or available in the data.
- business caseuses · A business case is a justification for undertaking an initiative that describes its expected value, benefits, costs, risks, and feasibility.
- decision analysisused by · Decision analysis is the systematic evaluation of alternatives, information, preferences, uncertain states, and consequences to guide a choice.
Questions about sensitivity analysis
What is sensitivity analysis?
Sensitivity analysis is a method for examining how changes in inputs, parameters, assumptions, or conditions affect analytical results or model behavior.
Which Manning books explain sensitivity analysis?
Build AI Drug Discovery Pipelines and Applied Reinforcement Learning.
What kinds of sensitivity analysis are there?
Kinds of sensitivity analysis include global sensitivity analysis, refutation and layer sensitivity analysis.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.

