Globally Optimal Design of a Distributed Scalar Quantizer for Linear Classification
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
This work is concerned with the design of a distributed scalar quantizer (DSQ) with two encoders, for linear classification. The objective of the optimization is to minimize the classification error of the classifier applied to the quantized inputs in the training sequence with respect to the classifier applied on unquantized inputs. We prove that the optimal DSQ design problem is equivalent to a minimum weight path problem with some constraints on the number and types of edges in a certain weighted directed acyclic graph. Further, we propose a solution algorithm with time complexity <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$O(K_{1}K_{2}N^{4})$</tex> , where <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$N$</tex> is the size of the training sequence while <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$K_{1}$</tex> and <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$K_{2}$</tex> are the numbers of cells of the two encoders, respectively. In addition, we develop faster design algorithms for the equal-rate case (i.e., <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$K_{1}=K_{2}=K$</tex> ). Specifically, when the training sequence is symmetric, we prove that there exists an optimal DSQ where the thresholds of the encoders' partitions are interleaved. By leveraging this property and the symmetry of the training sequence, we propose a <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$O(KN^{2})$</tex> time solution algorithm. For the case when the training sequence is not symmetric, we propose an algorithm with the same time complexity that minimizes an upper bound on the misclassification ratio. Experimental results prove the considerable superiority of the proposed approaches in comparison with prior work in both symmetric and asymmetric scenarios.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it