Library / Mathematical Modeling DPF
Jump to passage
In this reading

Link to current text

Published source confirmed at last check

Source changed 2026-10-03 08:25:59 UTC · snapshot created 2026-10-03 08:26:43 UTC · last check 2026-10-03 10:05:06 UTC

Table of Contents

Public units

UnitTitleUse
ReadmeMathematical Modeling - ReadmeFollow worked connections between modeling methods.
PrefaceMathematical Modeling - PrefaceUnderstand the connected methods, their rationale, sources and limits.

A. Formulate the subject question

§ID & TitleStatusKeywords & Search QueriesDependencies
1MMP.10 - Construct and Revise a Mathematical Constraint FormulationUsable, evolvingobjects; representation; domains; constraints; recovery; multiplicity; optimization. How can the objects and conditions of a working question be expressed so that mathematical answers still refer to the intended possibilities?B.5.FM for initial formulation; MATH.1/.5 for construction and representation; C.29.1 for transferring results; MMP.11 for unknown relations.
2MMP.11 - Construct a Mathematical Model Family from Known RelationsUsable, evolvingunknown relation; family; grounded structure; parameterization; completeness; identifiability. How can an unknown contribution vary without losing known relations, and which remaining ambiguity changes the requested consequence?MMP.10 for admissible representation; A.3.3.TR for state; C.16.IR for observation-based inference; C.28.MR for intervention; MMP.9 for reduction.
3MMP.8 - Formulate Information-Dependent Choices MathematicallyUsable, evolvingavailable information; action; policy; decision timing; objective; conditional requirement; average performance. What can be chosen with information available in time, and which conditions must that choice satisfy?MMP.7 for reports; A.3.3.TR for change; MMP.10 for constraints; ME for changing the corresponding way of working.
4MMP.8.SD - Construct a Sequential Decision Model from Information and ConsequencesUsable, evolvingsequential decision; sufficient state; belief; policy; transition; accumulated consequence; continuation. What information must a continuing instruction retain, and how do present choices affect later possibilities and results?MMP.8 for available choices; A.3.3.PI for predictive state and belief; MMP.7 for records; CMP.3/.5/.9 for policy computation.

B. Infer, distinguish and revise

§ID & TitleStatusKeywords & Search QueriesDependencies
1MMP.7 - Construct a Probability Model of the Recorded DataUsable, evolvingobservation; recorded data; selection; missing records; censoring; noise; probability law. What distribution does the recording procedure produce from the modeled possibilities?C.16.IR for interpretation and inference; MMP.11 for unknown relations; MMP.8 when the record informs a choice.
2MMP.12 - Formulate an Inverse Problem and Choose Whether and How to Use RegularizationUsable, evolvinginverse problem; identifiability; conditioning; regularization; restriction; penalty; target. Which unknowns can the records resolve, and what improvement and loss does a stable reconstruction introduce?C.16.IR for compatible cases; MMP.11/.7 for the forward and recording relations; MMP.13 for inferential uncertainty; CMP for solving the formulated problem.
3MMP.13 - Infer Unknowns Statistically under a Stated Observation ModelUsable, evolvingestimator; confidence coverage; prior; posterior; predictive uncertainty; dependence; target propagation. Which inferential construction supports the statement the receiver needs?MMP.7 for the record law; MMP.12/C.16.IR for ambiguity; CMP.8/.9 for numerical computation; MMP.14/.8 for prediction checks and decisions.
4MMP.14 - Find and Repair a Mathematical Model’s Failed PredictionsUsable, evolvingmodel criticism; predictive comparison; residual; conditional response; selection; repair; held-out prediction. Which discrepancy changes the use, what assumption can repair it, and what consequence must be recalculated?MMP.7/.11/.13 for comparable predictions; B.5.RR for affected revision; CMP.8/.9 for numerical error; C.11.DUA for worthwhile further checking.
5MMP.15 - Identify an Intervention Effect from Available DataUsable, evolvingcausal identification; intervention; adjustment; mediation; population transfer; support; partial identification. Which intervention consequence follows from the available laws and causal assumptions?C.28.MR for intervention meaning; MMP.7/.13 for recording and estimation; C.16.IR for compatible possibilities; MMP.16 for worthwhile distinguishing observations.
6MMP.19 - Construct and Bound a Mathematical Counterfactual Comparison under a Causal ModelStablecounterfactual; same case; factual inference; common input; joint response; pathway; bounds. What would happen to the same case under another action, and what does the available information determine?C.28.MR for changed mechanisms; MMP.7/.13 for factual records and inference; MMP.15 for identification; MMP.10/.16 for bounds and useful observations.
7MMP.16 - Design Observations to Separate Mathematical Model AlternativesUsable, evolvingobservation design; discrimination; nuisance; recording law; value of information; feasible design. Which obtainable observation separates a consequential ambiguity, and can its benefit repay its burden?MMP.7/.11/.12/.13 for predicted record laws; MMP.8/.8.SD for decisions; C.11.DUA for worthwhile further work; subject methods for realization.

C. Change the model while retaining its use

§ID & TitleStatusKeywords & Search QueriesDependencies
1MMP.9 - Derive a Reduced Mathematical Model of State EvolutionUsable, evolvingreduction; hidden state; closure; memory; approximation; error; sufficient bound. What contribution is left by eliminated detail, and which replacement preserves the consequence needed?A.3.3.TR for evolution; C.29.1 for preserved structure; MMP.11 for a replacement family; B.5.RR for affected revision.
2MMP.17 - Construct a Surrogate for Selected Model ResponsesUsable, evolvingsurrogate; selected response; input region; approximation; correction; tail; refinement. Which cheaper construction can supply the model response the next operation actually needs?MMP.11 for retained structure; CMP.7/.6 for learning and updates; MMP.14 for consequential mismatch; MMP.18 for coupled use.
3MMP.18 - Couple Models with Compatible Exchanges and ScalesUsable, evolvingcoupling; interface; scale; conservation; closure; shared uncertainty; double counting. How can component models exchange quantities and information without changing the required joint answer?A.3.3.TR and MMP.10 for joint conditions; MATH.18 for retained consequences; MMP.9/.17 for missing contributions; CMP.8/.14 for interacting computation.