Neural Networks and the Bias/Variance Dilemma.¶
Geman, S., Bienenstock, E., & Doursat, R. (1992). Neural Networks and the Bias/Variance Dilemma. Neural Computation, 4(1), 1-58.
Cited by¶
7 citations across 7 artifacts.
Each citation links to the sentence it supports in the citing article.
Primes¶
- Bias
- It separates two components of total error (the signed, reproducible part and the unsigned, averaging part) and names the part that carries a direction, a decomposition Geman, Bienenstock, and Doursat (1992) made canonical in their analysis of the bias/variance dilemma for learning systems.
This sourceCanonical decomposition of a learning system's total error into bias and variance components; grounds the cross-domain inference that aggregation can rescue an unbiased noisy process but never a biased one, recurring in polling, ensembles, sensor fusion, and forecasting.
Supported in partVerified against the publisher's abstract
The source is the canonical bias/variance-dilemma paper, but only an abstract was obtained and it never states the error decomposition, so the claim's signed/unsigned components remain unverified.
“Feedforward neural networks trained by error backpropagation are examples of nonparametric regression estimators.”
- It separates two components of total error (the signed, reproducible part and the unsigned, averaging part) and names the part that carries a direction, a decomposition Geman, Bienenstock, and Doursat (1992) made canonical in their analysis of the bias/variance dilemma for learning systems.
- Conjugate-Observable Complementarity
- In machine learning, the fit–generalization (bias–variance) trade-off is a conjugate pair on one model: driving training error toward zero (sharp fit) inflates variance and degrades generalization, and the two cannot both be made arbitrarily small on unknown data — regularization chooses a frontier point rather than abolishing the floor.
This sourceFormalizes the bias–variance decomposition as a conjugate trade-off in statistical learning, with a floored expected error.
- In machine learning, the fit–generalization (bias–variance) trade-off is a conjugate pair on one model: driving training error toward zero (sharp fit) inflates variance and degrades generalization, and the two cannot both be made arbitrarily small on unknown data — regularization chooses a frontier point rather than abolishing the floor.
- Grain of Analysis
- Qualitative coding: the coding grain (sentence, paragraph, episode, theme) determines whether patterns survive synthesis; overcoding fragments meaning, undercoding loses warranted distinctions. Quantitative model fitting: model complexity sets the grain at which a model resolves the data; overfitting resolves below the signal level, underfitting above it, and the bias-variance trade-off is the standard framing.
Supported in partVerified against the publisher's abstract
Title and abstract place this work in the bias/variance and model-complexity frame, but the readable text says nothing about coding grain or about resolving below the signal level.
“We present a tutorial on nonparametric inference and its relation to neural networks, and we use the statistical viewpoint to highlight strengths and weaknesses of neural models.”
- Qualitative coding: the coding grain (sentence, paragraph, episode, theme) determines whether patterns survive synthesis; overcoding fragments meaning, undercoding loses warranted distinctions. Quantitative model fitting: model complexity sets the grain at which a model resolves the data; overfitting resolves below the signal level, underfitting above it, and the bias-variance trade-off is the standard framing.
- No Free Lunch Theorem
- NFL is a conservation law over the space of problems constraining how any procedure performs averaged across all instances — a meta-claim about methods, not a method. Not the bias-variance trade-off alone. See Bias: that trade-off is the special case of NFL for estimation under squared error.
This sourceStandard reference for the bias-variance trade-off, the estimation-under-squared-error special case of the no-free-lunch conservation result.
Supported in partVerified against the publisher's abstract
Standard reference for the bias/variance trade-off in squared-error estimation; its abstract never mentions the no-free-lunch theorem, so the NFL special-case link is not carried here.
“We present a tutorial on nonparametric inference and its relation to neural networks, and we use the statistical viewpoint to highlight strengths and weaknesses of neural models.”
- NFL is a conservation law over the space of problems constraining how any procedure performs averaged across all instances — a meta-claim about methods, not a method. Not the bias-variance trade-off alone. See Bias: that trade-off is the special case of NFL for estimation under squared error.
- Overfitting
- The essential commitment is relational: overfitting is not a property of the model alone or the data alone, but of the model-data-target relationship, diagnosed by the gap between in-sample and out-of-sample performance
This sourceCanonical decomposition of a learning system's total error into bias and variance components; grounds the cross-domain inference that aggregation can rescue an unbiased noisy process but never a biased one, recurring in polling, ensembles, sensor fusion, and forecasting.
Supported in partVerified against the publisher's abstract
The canonical bias/variance paper, but held only as an abstract (the DOI fetch returned 403): its title and topic match the bias/variance framing, while the abstract itself states no decomposition of error into bias and variance.
“Neural Networks and the Bias/Variance Dilemma”
- The essential commitment is relational: overfitting is not a property of the model alone or the data alone, but of the model-data-target relationship, diagnosed by the gap between in-sample and out-of-sample performance
- Trade-offs
Mechanisms¶
- Complexity or Regularization Review
- The principle it enforces is the bias–variance trade-off: past a point, added flexibility reduces error on seen data while increasing it on unseen data.
This sourceDevelops the bias–variance dilemma in which increased model flexibility can lower fitting bias while raising variance and prediction error on new data.
- The principle it enforces is the bias–variance trade-off: past a point, added flexibility reduces error on seen data while increasing it on unseen data.
Verification¶
Does it exist? Confirmed. This work's DOI resolves to a registered record, which fixes its identity. That is all it fixes.
Does it back the claim? Read against the text for 4 of 7 citations: 4 supported in part. Each verdict is shown under its citation below, with what in the work backs the sentence.
Support is checked per citation rather than per work — the same source can be cited soundly in one article and wrongly in another. Per-citation recording began recently, so a citation with no recorded check is a gap in the record rather than evidence it went unchecked.
See how references were verified.
Registry ID ref:9ef11aedda7f · see in the full table