pith. sign in

arxiv: 2604.06451 · v1 · submitted 2026-04-07 · 💻 cs.LG

Quality-preserving Model for Electronics Production Quality Tests Reduction

Noufa Haneefa , Teddy Lazebnik , Einav Peretz-Andersson This is my paper

Pith reviewed 2026-05-10 19:41 UTC · model grok-4.3

classification 💻 cs.LG
keywords adaptive test selectionelectronics manufacturingquality controlmulti-armed banditconcept drifttest reductionprinted circuit boardThompson sampling
0
0 comments X

The pith

An adaptive framework using offline set cover and online Thompson sampling eliminates defect escapes in electronics testing while cutting test times.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper demonstrates that fixed test plans in high-volume electronics production become inefficient as failure patterns evolve, leading to unnecessary costs without guaranteed quality. It proposes a hybrid system that first constructs reduced diagnostic test subsets offline via greedy set cover to minimize cost at zero escape risk, then deploys a Thompson-sampling bandit online to switch between those subsets and full coverage according to a rolling process-stability signal. Evaluated on 28,000 real board runs across Functional Circuit Test and End-of-Line stages with actual concept drift, the adaptive policy achieved zero escapes by reverting during instability, whereas static reduced plans allowed 110 and 8 escapes respectively. This matters for manufacturers because it offers measurable cost and time savings without compromising outgoing quality when conditions change.

Core claim

The adaptive test-selection framework, built from offline minimum-cost diagnostic subsets via greedy set cover and an online Thompson-sampling multi-armed bandit controlled by a rolling process-stability signal, produced zero escaped defects under temporal validation with real concept drift on two printed circuit board assembly stages, while static reduction plans produced 110 escapes in Functional Circuit Test and 8 in End-of-Line testing.

What carries the argument

The Thompson-sampling multi-armed bandit that switches between full and reduced test plans using a rolling process-stability signal.

If this is right

  • Offline construction yields zero-escape reduced plans that cut test time by 18.78 percent in Functional Circuit Test and 91.57 percent in End-of-Line testing.
  • The online bandit policy maintains zero escapes by reverting to fuller coverage precisely when instability is detected.
  • The combined offline-plus-online approach supplies a concrete route to adaptive test planning that balances cost reduction against escape risk across production stages.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same stability-triggered switching logic could be tested on other manufacturing lines that face gradual shifts in defect distributions.
  • Integrating the bandit decision layer with existing factory data pipelines would allow real-time test adjustment without manual intervention.
  • Extending the offline subset construction to include multiple cost objectives (time, equipment wear, operator effort) could further improve practical adoption.

Load-bearing premise

The rolling process-stability signal reliably detects periods when reduced testing would allow defect escapes.

What would settle it

A production run where the stability signal indicates stability, the system uses a reduced test plan, and one or more defects still escape detection.

Figures

Figures reproduced from arXiv: 2604.06451 by Einav Peretz-Andersson, Noufa Haneefa, Teddy Lazebnik.

Figure 1
Figure 1. Figure 1: illustrates the overall test flow evaluated in this work [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: A schematic view of the multi-arm bandit algorithms and their production tests [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Schematic flow of the proposed framework. [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Cost-risk Pareto fronts for the two test suites. The grey dotted line shows a logarithmic [PITH_FULL_IMAGE:figures/full_fig_p013_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Rolling Cred selection fraction during training for the Thompson Sampling agent on the FCT and EOL datasets [PITH_FULL_IMAGE:figures/full_fig_p013_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Rolling pass rate and Cred selection behavior during validation for the FCT and EOL datasets. 5 Discussion In this study, we proposed an MAB-based model for quality-preserving production test reduction. We evaluate the proposed model on a 2-step test set from a real-world production line of electronics. The results demonstrate that the value of the proposed framework lies not only in reducing test effort, … view at source ↗
read the original abstract

Manufacturing test flows in high-volume electronics production are typically fixed during product development and executed unchanged on every unit, even as failure patterns and process conditions evolve. This protects quality, but it also imposes unnecessary test cost, while existing data-driven methods mostly optimize static test subsets and neither adapt online to changing defect distributions nor explicitly control escape risk. In this study, we present an adaptive test-selection framework that combines offline minimum-cost diagnostic subset construction using greedy set cover with an online Thompson-sampling multi-armed bandit that switches between full and reduced test plans using a rolling process-stability signal. We evaluate the framework on two printed circuit board assembly stages-Functional Circuit Test and End-of-Line test-covering 28,000 board runs. Offline analysis identified zero-escape reduced plans that cut test time by 18.78% in Functional Circuit Test and 91.57\% in End-of-Line testing. Under temporal validation with real concept drift, static reduction produced 110 escaped defects in Functional Circuit Test and 8 in End-of-Line, whereas the adaptive policy reduced escapes to zero by reverting to fuller coverage when instability emerged in practice. These results show that online learning can preserve manufacturing quality while reducing test burden, offering a practical route to adaptive test planning across production domains, and offering both economic and logistics improvement for companies.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper proposes an adaptive test-selection framework for high-volume electronics manufacturing that first uses offline greedy set cover to construct minimum-cost reduced diagnostic subsets and then deploys an online Thompson-sampling multi-armed bandit. The bandit switches between full and reduced test plans on the basis of a rolling process-stability signal. On a temporal split of 28,000 real PCB boards from Functional Circuit Test and End-of-Line stages, static reduced plans produced 110 and 8 escapes respectively, while the adaptive policy achieved zero escapes by reverting to fuller coverage when the stability signal indicated drift.

Significance. If the central result holds, the work supplies a practical, data-driven route to lowering test cost (reported 18.78 % and 91.57 % reductions) while explicitly controlling escape risk under real concept drift—an improvement over purely static reduction methods that dominate current practice. The combination of offline set-cover optimization with an online bandit that uses an external stability signal, together with temporal validation on production data, is a concrete strength that could be adopted in other manufacturing domains.

major comments (2)
  1. [Online Thompson-sampling component (Section 3.2)] The zero-escape claim for the adaptive policy rests entirely on the rolling process-stability signal correctly detecting instability and triggering reversion to full coverage. The manuscript provides no definition of this signal (input metrics, window length, statistic, or threshold) nor any ablation or sensitivity analysis of its parameters; without these details it is impossible to determine whether the reported advantage is a general mechanism or an artifact of tuning to the specific drift patterns in the 28,000-board temporal split.
  2. [Evaluation under temporal validation (Section 4)] Table 2 (or equivalent results table) reports escape counts of 110/8 for static reduction versus 0 for the adaptive policy, yet no statistical test, confidence interval, or comparison against alternative adaptive baselines (e.g., other bandit formulations or simple threshold rules) is supplied. This weakens the strength of the cross-policy claim.
minor comments (2)
  1. [Abstract] The abstract states '91.57% in End-of-Line testing' but the percentage sign is inconsistently formatted; standardize notation throughout.
  2. [Offline subset construction (Section 3.1)] The description of the offline greedy set cover algorithm would benefit from an explicit statement of the cost function and the stopping criterion used to guarantee zero escapes on the training data.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive and detailed feedback. The comments identify important gaps in methodological description and evaluation rigor that we will address in revision. Below we respond point-by-point to the major comments.

read point-by-point responses

  1. Referee: [Online Thompson-sampling component (Section 3.2)] The zero-escape claim for the adaptive policy rests entirely on the rolling process-stability signal correctly detecting instability and triggering reversion to full coverage. The manuscript provides no definition of this signal (input metrics, window length, statistic, or threshold) nor any ablation or sensitivity analysis of its parameters; without these details it is impossible to determine whether the reported advantage is a general mechanism or an artifact of tuning to the specific drift patterns in the 28,000-board temporal split.

    Authors: We acknowledge that the description of the rolling process-stability signal in Section 3.2 is insufficiently detailed. In the revised manuscript we will explicitly specify the signal definition, including the input metrics (rolling defect rate and variance computed over a sliding window), the exact window length, the statistic employed, and the threshold used to trigger reversion to full coverage. We will also add a sensitivity analysis and ablation study that varies these parameters across a range of plausible values and reports the resulting escape counts and test-time reductions. This will demonstrate that the zero-escape outcome is robust rather than an artifact of tuning to the particular drift patterns in the 28k-board dataset. revision: yes

  2. Referee: [Evaluation under temporal validation (Section 4)] Table 2 (or equivalent results table) reports escape counts of 110/8 for static reduction versus 0 for the adaptive policy, yet no statistical test, confidence interval, or comparison against alternative adaptive baselines (e.g., other bandit formulations or simple threshold rules) is supplied. This weakens the strength of the cross-policy claim.

    Authors: We agree that the current presentation of results would be strengthened by statistical support and additional baselines. In the revised Section 4 we will augment the escape-count comparisons with appropriate statistical tests (e.g., McNemar’s test for paired binary outcomes) together with confidence intervals or bootstrap estimates for the escape rates. We will also report results for two alternative adaptive policies—an epsilon-greedy bandit and a simple threshold rule on the same stability signal—to provide direct context for the Thompson-sampling policy. These additions will allow readers to assess the relative contribution of the bandit formulation more rigorously. revision: yes

Circularity Check

0 steps flagged

No circularity: derivation uses standard algorithms and external signal without self-referential reduction.

full rationale

The framework applies off-the-shelf greedy set cover for offline minimum-cost subset selection and Thompson sampling for the online multi-armed bandit, with decisions driven by a rolling process-stability signal treated as an external input. The reported results compare static versus adaptive policies on a temporal split of 28,000 boards under real drift; zero escapes under the adaptive policy follow from the policy's switching behavior on observed instability rather than from any parameter fitted to the escape count itself or from a self-citation chain. No equation or step equates the outcome to its own inputs by construction, and the cited methods are independent of the target result.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on the assumption that the stability signal correlates with escape risk and that the offline subset is truly zero-escape; no explicit free parameters or invented entities are named in the abstract.

axioms (1)
  • domain assumption Rolling process-stability signal correlates with defect escape risk
    Invoked to justify switching between reduced and full plans without escapes.

pith-pipeline@v0.9.0 · 5536 in / 1174 out tokens · 74053 ms · 2026-05-10T19:41:11.069246+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

78 extracted references · 78 canonical work pages · 1 internal anchor

  1. [1]

    Guide to good manufacturing practice for medicinal products.Annexes PE, pages 009–6, 2009

    Pharmaceutical Inspection Co-operation Scheme. Guide to good manufacturing practice for medicinal products.Annexes PE, pages 009–6, 2009

    work page 2009

  2. [2]

    Rethinking pharmaceutical industry with quality by design: application in research, development, manufacturing, and quality assurance.The AAPS journal, 27(4):96, 2025

    Joana Galvão Duarte, Maria Galvão Duarte, Ana Paula Piedade, and Filipa Mascarenhas- Melo. Rethinking pharmaceutical industry with quality by design: application in research, development, manufacturing, and quality assurance.The AAPS journal, 27(4):96, 2025

    work page 2025

  3. [3]

    Enhancing production efficiency through digital twin simulation scheduling.Applied Sciences, 15(7):3637, 2025

    Patrik Grznár, Ladislav Papánek, Milan Marčan, Martin Krajčovič, Ivan Antoniuk, Štefan Mozol, and Lucia Mozolová. Enhancing production efficiency through digital twin simulation scheduling.Applied Sciences, 15(7):3637, 2025

    work page 2025

  4. [4]

    A review analysis of cost of quality, cost of poor quality, and hidden quality cost.The Journal of Technology Studies, 50:1–13, 06 2025

    Jonathan Walston, M.Affan Badar, Christopher Kluse, and Riem Rostom. A review analysis of cost of quality, cost of poor quality, and hidden quality cost.The Journal of Technology Studies, 50:1–13, 06 2025

    work page 2025

  5. [5]

    Zelin Tong, Jingdan Feng, and Fang Liu. Understanding damage to and reparation of brand trust: a closer look at image congruity in the context of negative publicity.Journal of Product & Brand Management, 32(1):157–170, 2023

    work page 2023

  6. [6]

    Impact of product quality on customer satisfaction: A systematic literature review

    Anuj Bhowmick and Arumugam Seetharaman. Impact of product quality on customer satisfaction: A systematic literature review. InProceedings of the 2023 7th International Conference on Virtual and Augmented Reality Simulations, pages 93–99, 2023

    work page 2023

  7. [7]

    Venkata Charan Kantumchu, Syed Quadir Moinuddin, Ashok Kumar Dewangan, and Mu- ralimohan Cheepu.Quality Assurance and Control in Welding and Additive Manufacturing, pages 245–261. 2024

    work page 2024

  8. [8]

    Bridging the gap between high and low-volume production through enhancement of integrative capabilities.Procedia Manufacturing, 5:26–40, 2016

    Elisabeth Lervåg Synnes and Torgeir Welo. Bridging the gap between high and low-volume production through enhancement of integrative capabilities.Procedia Manufacturing, 5:26–40, 2016

    work page 2016

  9. [9]

    A signature test framework for rapid production testing of rf circuits

    Ramakrishna Voorakaranam, Sasikumar Cherubal, and Abhijit Chatterjee. A signature test framework for rapid production testing of rf circuits. InProceedings 2002 Design, Automation and Test in Europe Conference and Exhibition, pages 186–191. IEEE, 2002. 18

    work page 2002

  10. [10]

    Minimizing production test time to detect faults in analog circuits.IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 13(6):796–813, 1994

    Linda Milor and Alberto L Sangiovanni-Vincentelli. Minimizing production test time to detect faults in analog circuits.IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 13(6):796–813, 1994

    work page 1994

  11. [11]

    Jon M. Quigley. Five approaches to product testing.IEEE Reliability Magazine, 1(1):30–36, 2024

    work page 2024

  12. [12]

    Chapter 1 - best practices in mission-assured, mission-critical, and safety-critical systems

    Kim Fowler. Chapter 1 - best practices in mission-assured, mission-critical, and safety-critical systems. In Kim Fowler, editor,Mission-Critical and Safety-Critical Systems Handbook, pages 1–82. Newnes, Boston, 2010

    work page 2010

  13. [13]

    Performance evaluation of multi-stage manufacturing systems operating under feedback and feedforward quality control loops.CIRP Annals, 73(1):349–352, 2024

    Maria Chiara Magnanini, Ozan Demir, Marcello Colledani, and Tullio Tolio. Performance evaluation of multi-stage manufacturing systems operating under feedback and feedforward quality control loops.CIRP Annals, 73(1):349–352, 2024

    work page 2024

  14. [14]

    Testing in the incremental design and development of complex products.Research in Engineering Design, 30, 04 2019

    Khadija Tahera, David Wynn, Chris Earl, and Claudia Eckert. Testing in the incremental design and development of complex products.Research in Engineering Design, 30, 04 2019

    work page 2019

  15. [15]

    Design, testing, and troubleshooting of industrial equipment: A systematic review of integration techniques for us manufacturing plants

    Adar Chowdhury and Md Nuruzzaman. Design, testing, and troubleshooting of industrial equipment: A systematic review of integration techniques for us manufacturing plants. Review of Applied Science and Technology, 2(01):53–84, 2023

    work page 2023

  16. [16]

    Embodied intelligence in manufacturing: leveraging large language models for autonomous industrial robotics.Journal of Intelligent Manufacturing, 36(2):1141–1157, 2025

    Haolin Fan, Xuan Liu, Jerry Ying Hsi Fuh, Wen Feng Lu, and Bingbing Li. Embodied intelligence in manufacturing: leveraging large language models for autonomous industrial robotics.Journal of Intelligent Manufacturing, 36(2):1141–1157, 2025

    work page 2025

  17. [17]

    Model- based test design of product lines: Raising test design to the product line level

    Hartmut Lackner, Martin Thomas, Florian Wartenberg, and Stephan Weißleder. Model- based test design of product lines: Raising test design to the product line level. In2014 IEEE Seventh International Conference on Software Testing, Verification and Validation, pages 51–60, 2014

    work page 2014

  18. [18]

    The role of testing in the engineering product development process

    Khadija Tahera, Christopher Earl, and Claudia Eckert. The role of testing in the engineering product development process. 05 2012

    work page 2012

  19. [19]

    A comprehensive scan test cost model to optimize the production of very large socs.IEEE Transactions on Computers, 74(4):1278–1292, 2025

    Giusy Iaria, Paolo Bernardi, Claudia Bertani, Lorenzo Cardone, Giuseppe Garozzo, and Vincenzo Tancorre. A comprehensive scan test cost model to optimize the production of very large socs.IEEE Transactions on Computers, 74(4):1278–1292, 2025

    work page 2025

  20. [20]

    Measurements for quality control in biomedical 3-d printing: Standards, gaps, and missing protocols.IEEE Sensors Reviews, 3, 2026

    Marta Cecchitelli, Giorgia Fiori, Jan Galo, Salvatore Andrea Sciuto, and Andrea Scorza. Measurements for quality control in biomedical 3-d printing: Standards, gaps, and missing protocols.IEEE Sensors Reviews, 3, 2026

    work page 2026

  21. [21]

    Artificial intelligence for smart manufacturing: Methods and applications

    Kim Phuc TRAN. Artificial intelligence for smart manufacturing: Methods and applications. Sensors, 21, 08 2021

    work page 2021

  22. [22]

    Development of an ai model adaptation strategy based on multi-resolution concept drift detection: A case study in virtual metrology for smart manufacturing

    Dong-Hyuk Yang, Hyeongjun Lee, and Yong-Shin Kang. Development of an ai model adaptation strategy based on multi-resolution concept drift detection: A case study in virtual metrology for smart manufacturing. In2025 IEEE Annual Congress on Artificial Intelligence of Things (AIoT), pages 335–341, 2025

    work page 2025

  23. [23]

    Ai-driven model-retraining architecture to sustain operational accuracy in data-drifting environments

    Hoo Meng Wong and Sundresan Perumal. Ai-driven model-retraining architecture to sustain operational accuracy in data-drifting environments. In2025 IEEE Symposium on Wireless Technology & Applications (ISWTA), pages 1–6, 2025

    work page 2025

  24. [24]

    Manufacturing as a data-driven practice: Methodologies, technologies, and tools.Proceedings of the IEEE, 109(4):399–422, 2021

    Tania Cerquitelli, Daniele Jahier Pagliari, Andrea Calimera, Lorenzo Bottaccioli, Edoardo Patti, Andrea Acquaviva, and Massimo Poncino. Manufacturing as a data-driven practice: Methodologies, technologies, and tools.Proceedings of the IEEE, 109(4):399–422, 2021. 19

    work page 2021

  25. [25]

    A review of artificial intelligence applications in manufacturing operations.Journal of Advanced Manufacturing and Processing, 5(3):e10159, 2023

    Siby Jose Plathottam, Arin Rzonca, Rishi Lakhnori, and Chukwunwike O Iloeje. A review of artificial intelligence applications in manufacturing operations.Journal of Advanced Manufacturing and Processing, 5(3):e10159, 2023

    work page 2023

  26. [26]

    Ai-driven quality control in pcb manufacturing: Enhancing produc- tion efficiency and precision.Valley International Journal Digital Library, 12(10):1549–1564, 2024

    Harshitkumar Ghelani. Ai-driven quality control in pcb manufacturing: Enhancing produc- tion efficiency and precision.Valley International Journal Digital Library, 12(10):1549–1564, 2024

    work page 2024

  27. [27]

    Advanced ai technologies for defect prevention and yield optimization in pcb manufacturing.International Journal Of Engineering And Computer Science, 13(10), 2024

    Harshitkumar Ghelani. Advanced ai technologies for defect prevention and yield optimization in pcb manufacturing.International Journal Of Engineering And Computer Science, 13(10), 2024

    work page 2024

  28. [28]

    Bathrinath, Praveen Vundrajavarapu, R

    M Anusuya, Kavitha P, S. Bathrinath, Praveen Vundrajavarapu, R. Bharath Kumar, and M. Sakthivel. Automation of test and measurement in electronics manufacturing through ai. In2024 International Conference on Science Technology Engineering and Management (ICSTEM), pages 1–6, 2024

    work page 2024

  29. [29]

    Anurag Vijay Agrawal, Kommisetti Murthy Raju, Aparna P, Gade Sravya, A Chan- drashekhar, and J. Ramya. Ai-driven test and measurement automation in electronics manufacturing. In2024 Ninth International Conference on Science Technology Engineering and Mathematics (ICONSTEM), pages 1–6, 2024

    work page 2024

  30. [30]

    Low test cost adaptive testing method for high yield ic products.Integration, 103:102401, 2025

    Yuqi Pan, Huaguo Liang, Junming Li, Zhengfeng Huang, Maoxiang Yi, and Yingchun Lu. Low test cost adaptive testing method for high yield ic products.Integration, 103:102401, 2025

    work page 2025

  31. [31]

    Recent advances of artificial intelligence in manufacturing industrial sectors: A review.International journal of precision engineering and manufacturing, 23(1):111–129, 2022

    Sung Wook Kim, Jun Ho Kong, Sang Won Lee, and Seungchul Lee. Recent advances of artificial intelligence in manufacturing industrial sectors: A review.International journal of precision engineering and manufacturing, 23(1):111–129, 2022

    work page 2022

  32. [32]

    A surprisal- based greedy heuristic for the set covering problem, 06 2023

    Tommaso Adamo, Gianpaolo Ghiani, Emanuela Guerriero, and Deborah Pareo. A surprisal- based greedy heuristic for the set covering problem, 06 2023

    work page 2023

  33. [33]

    Daniel Russo, Benjamin Roy, Abbas Kazerouni, Ian Osband, and Zheng Wen.A Tutorial on Thompson Sampling. 2018

    work page 2018

  34. [34]

    Multi-armed bandit test case prioritiza- tion in continuous integration environments: A trade-off analysis

    Jackson Antonio do Prado Lima and Silvia Vergilio. Multi-armed bandit test case prioritiza- tion in continuous integration environments: A trade-off analysis. 10 2020

    work page 2020

  35. [35]

    Boumen, I

    R. Boumen, I. S. M. de Jong, J. W. H. Vermunt, J. M. van de Mortel-Fronczak, and J. E. Rooda. Test sequencing in complex manufacturing systems.IEEE Transactions on Systems, Man, and Cybernetics - Part A: Systems and Humans, 38(1):25–37, 2008

    work page 2008

  36. [36]

    Optimal sampling and decision-making strategies in multi-stage manufacturing processes

    Xu Wang and Zhiyan Yang. Optimal sampling and decision-making strategies in multi-stage manufacturing processes. In2025 10th International Conference on Cloud Computing and Big Data Analytics (ICCCBDA), pages 658–663, 2025

    work page 2025

  37. [37]

    Ender Yilmaz, Sule Ozev, and Kenneth M. Butler. Adaptive test flow for mixed-signal/rf circuits using learned information from device under test. In2010 IEEE International Test Conference, pages 1–10, Nov 2010

    work page 2010

  38. [38]

    defect level versus cost

    M. Scheffler, P.D. Franzon, and G. Troster. A "defect level versus cost" system tradeoff for electronics manufacturing.IEEE Transactions on Electronics Packaging Manufacturing, 27(1):67–76, 2004. 20

    work page 2004

  39. [39]

    Test cost-test quality modeling for adaptive test

    Bekir Zahit Demiray and Baris Arslan. Test cost-test quality modeling for adaptive test. In 2022 IEEE International Conference on Automation, Quality and Testing, Robotics (AQTR), pages 1–5, 2022

    work page 2022

  40. [40]

    Saxena, Edward J

    François-Fabien Ferhani, Nirmal R. Saxena, Edward J. McCluskey, and Phil Nigh. How many test patterns are useless? In26th IEEE VLSI Test Symposium (vts 2008), pages 23–28, 2008

    work page 2008

  41. [41]

    Benner and O

    S. Benner and O. Boroffice. Optimal production test times through adaptive test program- ming. InProceedings International Test Conference 2001 (Cat. No.01CH37260), pages 908–915, 2001

    work page 2001

  42. [42]

    Adam Hamrol, Agnieszka Kujawińska, and Mariusz Bożek. Quality inspection planning within a multistage manufacturing process based on the added value criterion.The Interna- tional Journal of Advanced Manufacturing Technology, 108:1–14, 05 2020

    work page 2020

  43. [43]

    An industry example to reduce the test time by optimizing data extration method

    Xiaofeng Liang and Deguang Zheng. An industry example to reduce the test time by optimizing data extration method. In2022 China Semiconductor Technology International Conference (CSTIC), pages 1–4, 2022

    work page 2022

  44. [44]

    Nikolay Petkov and Malinka Ivanova. Printed circuit board and printed circuit board assem- bly methods for testing and visual inspection: a review.Bulletin of Electrical Engineering and Informatics, 13:2566–2585, 08 2024

    work page 2024

  45. [45]

    Optimized lightweight pcb real-time defect detection algorithm

    Dahua Li, Ao Xu, and Xiao Yu. Optimized lightweight pcb real-time defect detection algorithm. In2023 IEEE 16th International Conference on Electronic Measurement & Instruments (ICEMI), pages 262–269, 2023

    work page 2023

  46. [46]

    Universal platform for pcb functional testing

    Marian Serban, Yuriy Vagapov, Zheng Chen, Reg Holme, and Sergey Lupin. Universal platform for pcb functional testing. pages 402–409, 09 2014

    work page 2014

  47. [47]

    Árpád Kis, Cosmin Ancuti, and Codruta O. Ancuti. Ats-pcb: An effective automated testing system for advanced driver assistance systems. In2019 International Symposium ELMAR, pages 215–218, 2019

    work page 2019

  48. [48]

    Developing computational intelligence for smart qualification testing of electronic products.IEEE Access, 8:16922–16933, 2020

    Mominul Ahsan, Stoyan Stoyanov, Chris Bailey, and Alhussein Albarbar. Developing computational intelligence for smart qualification testing of electronic products.IEEE Access, 8:16922–16933, 2020

    work page 2020

  49. [49]

    Machine learning algorithms for manufacturing quality assurance: A systematic review of performance metrics and applications.Array, 26:100393, 2025

    Ashfakul Karim Kausik, Adib Bin Rashid, Ramisha Fariha Baki, and Md Mifthahul Jannat Maktum. Machine learning algorithms for manufacturing quality assurance: A systematic review of performance metrics and applications.Array, 26:100393, 2025

    work page 2025

  50. [50]

    Adaptation of functional inspection test plan in a production line using a multi-agent system

    Nelson Rodrigues, Paulo Leitão, Matthias Foehr, Claudio Turrin, Arnaldo Pagani, and Roberto Decesari. Adaptation of functional inspection test plan in a production line using a multi-agent system. In2013 IEEE International Symposium on Industrial Electronics, pages 1–6, 2013

    work page 2013

  51. [51]

    Test-suite-based analog/rf test time reduction using canonical correlation.IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 35(12):2143–2147, 2016

    Ke Huang, Jian Wen, and Jim Willmore. Test-suite-based analog/rf test time reduction using canonical correlation.IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 35(12):2143–2147, 2016

    work page 2016

  52. [52]

    Machine learning classification algorithm for vlsi test cost reduction.Integration, 87:40–48, 2022

    Tai Song, Zhengfeng Huang, and Aibin Yan. Machine learning classification algorithm for vlsi test cost reduction.Integration, 87:40–48, 2022. 21

    work page 2022

  53. [53]

    Adaptive testing of compute-in-memory based cnns using probabilistic test acceptance limits

    Anurup Saha, Kwondo Ma, Chandramouli Amarnath, Moinuddin Qureshi, and Abhijit Chatterjee. Adaptive testing of compute-in-memory based cnns using probabilistic test acceptance limits. In2025 IEEE 31st International Symposium on On-Line Testing and Robust System Design (IOLTS), pages 1–7, 2025

    work page 2025

  54. [54]

    Adaptive test framework for achieving target test quality at minimal cost

    Baris Arslan and Alex Orailoglu. Adaptive test framework for achieving target test quality at minimal cost. In2011 Asian Test Symposium, pages 323–328, 2011

    work page 2011

  55. [55]

    Introduction to multi-armed bandits.Foundations and Trends®in Machine Learning, 12(1-2):1–286, 2019

    Aleksandrs Slivkins. Introduction to multi-armed bandits.Foundations and Trends®in Machine Learning, 12(1-2):1–286, 2019

    work page 2019

  56. [56]

    Multi-armed bandit algorithms and empirical evaluation

    Joannes Vermorel and Mehryar Mohri. Multi-armed bandit algorithms and empirical evaluation. InEuropean conference on machine learning, pages 437–448. Springer, 2005

    work page 2005

  57. [57]

    Multi-armed bandit problems

    Aditya Mahajan and Demosthenis Teneketzis. Multi-armed bandit problems. InFoundations and applications of sensor management, pages 121–151. Springer, 2008

    work page 2008

  58. [58]

    Teddy Lazebnik. Data-driven hospitals staff and resources allocation using agent-based sim- ulation and deep reinforcement learning.Engineering Applications of Artificial Intelligence, 126:106783, 2023

    work page 2023

  59. [59]

    Exploration– exploitation model of moth-inspired olfactory navigation.Journal of the Royal Society Interface, 21(216), 2024

    Teddy Lazebnik, Yiftach Golov, Roi Gurka, Ally Harari, and Alex Liberzon. Exploration– exploitation model of moth-inspired olfactory navigation.Journal of the Royal Society Interface, 21(216), 2024

    work page 2024

  60. [60]

    home venues

    TeddyLazebnik, ShirAviv-Reuven, andArielRosenfeld. Publishinginstincts: Anexploration- exploitation framework for studying academic publishing behavior and “home venues”. Journal of Informetrics, 19(3):101705, 2025

    work page 2025

  61. [61]

    Aggarwal

    Djallel Bouneffouf, Irina Rish, and Charu C. Aggarwal. Survey on applications of multi- armed and contextual bandits. In2020 IEEE Congress on Evolutionary Computation (CEC), pages 1–8, 2020

    work page 2020

  62. [62]

    A learning-based context-aware quality test system in B5G-aided advanced manufacturing.IEEE Transactions on Industrial Informatics, 19(2):1548–1558, 2023

    Sige Liu, Peng Cheng, Zhuo Chen, Kan Yu, Wei Xiang, Jun Li, Branka Vucetic, and Yonghui Li. A learning-based context-aware quality test system in B5G-aided advanced manufacturing.IEEE Transactions on Industrial Informatics, 19(2):1548–1558, 2023

    work page 2023

  63. [63]

    Prado Lima and Silvia Regina Vergilio

    Jackson A. Prado Lima and Silvia Regina Vergilio. A multi-armed bandit approach for test case prioritization in continuous integration environments.IEEE Transactions on Software Engineering, 48(2):453–465, 2022

    work page 2022

  64. [64]

    Survey on applications of multi-armed and contextual bandits

    Djallel Bouneffouf, Irina Rish, and Charu Aggarwal. Survey on applications of multi-armed and contextual bandits. In2020 IEEE Congress on Evolutionary Computation (CEC), pages 1–8, 2020

    work page 2020

  65. [65]

    The economical-ecological benefits of matching non-matching socks.arXiv preprint arXiv:2602.18221, 2026

    Teddy Lazebnik. The economical-ecological benefits of matching non-matching socks.arXiv preprint arXiv:2602.18221, 2026

    work page internal anchor Pith review arXiv 2026

  66. [66]

    Exploring the depths of multi-armed bandit algorithms: From theoretical foundations to modern applications.Applied and Computational Engineering, 68:183–191, 06 2024

    Haoyu Nie. Exploring the depths of multi-armed bandit algorithms: From theoretical foundations to modern applications.Applied and Computational Engineering, 68:183–191, 06 2024

    work page 2024

  67. [67]

    Adaptive policy optimization for product infant failure risk control in manufacturing processes with variable demands

    Xiao Han, Yuqi Cai, Anqi Zhang, Yao Zhu, Yihai He, and Rui Shi. Adaptive policy optimization for product infant failure risk control in manufacturing processes with variable demands. In2025 16th International Conference on Reliability, Maintainability and Safety (ICRMS), pages 376–381, 2025. 22

    work page 2025

  68. [68]

    A greedy heuristic for the set-covering problem.Math

    Vasek Chvátal. A greedy heuristic for the set-covering problem.Math. Oper. Res., 4:233–235, 1979

    work page 1979

  69. [69]

    A learning-based context-aware quality test system in b5g-aided advanced manufacturing

    Sige Liu, Peng Cheng, Zhuo Chen, Kan Yu, Wei Xiang, Jun Li, Branka Vucetic, and Yonghui Li. A learning-based context-aware quality test system in b5g-aided advanced manufacturing. IEEE Transactions on Industrial Informatics, 04 2022

    work page 2022

  70. [70]

    G. Burg, A. Nazábal, and C. Sutton. Wrangling messy csv files by detecting row and type patterns.Data Mining and Knowledge Discovery, 33, 11 2019

    work page 2019

  71. [71]

    A review on data normalization techniques.International Journal of Engineering Research and, V9, 07 2020

    Kalyani Sankpal. A review on data normalization techniques.International Journal of Engineering Research and, V9, 07 2020

    work page 2020

  72. [72]

    Advancing zero defect manufacturing: A state-of-the-art perspective and future research directions

    Daryl Powell, Maria Chiara Magnanini, Marcello Colledani, and Odd Myklebust. Advancing zero defect manufacturing: A state-of-the-art perspective and future research directions. Computers in Industry, 136:103596, 2022

    work page 2022

  73. [73]

    Digitalization platform for data-driven quality management in multi-stage manufacturing systems.Journal of Intelligent Manufacturing, 35(6):2699–2718, 2024

    Marc-André Filz, Jan Philipp Bosse, and Christoph Herrmann. Digitalization platform for data-driven quality management in multi-stage manufacturing systems.Journal of Intelligent Manufacturing, 35(6):2699–2718, 2024

    work page 2024

  74. [74]

    Adam Hamrol, Agnieszka Kujawińska, and Mariusz Bożek. Quality inspection planning within a multistage manufacturing process based on the added value criterion.The Interna- tional Journal of Advanced Manufacturing Technology, 108(5–6):1399–1412, 2020

    work page 2020

  75. [75]

    Cost of quality: Evaluating cost-quality trade-offs for inspection strategies of manufacturing processes.International Journal of Production Economics, 188:156–166, 2017

    Muhammad Arsalan Farooq, Randolph Kirchain, Henriqueta Novoa, and Antonio Araújo. Cost of quality: Evaluating cost-quality trade-offs for inspection strategies of manufacturing processes.International Journal of Production Economics, 188:156–166, 2017

    work page 2017

  76. [76]

    Conservative bandits

    Yifan Wu, Roshan Shariff, Tor Lattimore, and Csaba Szepesvári. Conservative bandits. InProceedings of the 33rd International Conference on Machine Learning, volume 48 of Proceedings of Machine Learning Research, pages 1254–1262. PMLR, 2016

    work page 2016

  77. [77]

    Conservative contextual linear bandits

    Abbas Kazerouni, Mohammad Ghavamzadeh, Yasin Abbasi-Yadkori, and Benjamin Van Roy. Conservative contextual linear bandits. InAdvances in Neural Information Processing Systems 30, pages 3910–3919. Curran Associates, Inc., 2017

    work page 2017

  78. [78]

    Mostafa, and Ahmed El-assal

    Mohamed Ismail, Noha A. Mostafa, and Ahmed El-assal. Quality monitoring in multi- stage manufacturing systems by using machine learning techniques.Journal of Intelligent Manufacturing, 33:2471–2486, 2022. 23

    work page 2022