Hidden Technical Debt in Machine Learning Systems

Authors: D. Sculley, Gary Holt, Daniel Golovin, Eugene Davydov, Todd Phillips, Dietmar Ebner, Vinay Chaudhary, Michael Young, Jean-François Crespo, Dan Dennison

NeurIPS 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Theoretical This paper does not offer novel ML algorithms, but instead seeks to increase the community s awareness of the difficult tradeoffs that must be considered in practice over the long term. We focus on system-level interactions and interfaces as an area where ML technical debt may rapidly accumulate.
Researcher Affiliation Industry D. Sculley, Gary Holt, Daniel Golovin, Eugene Davydov, Todd Phillips {dsculley,gholt,dgg,edavydov,toddphillips}@google.com Google, Inc. Dietmar Ebner, Vinay Chaudhary, Michael Young, Jean-Franc ois Crespo, Dan Dennison {ebner,vchaudhary,mwyoung,jfcrespo,dennison}@google.com Google, Inc.
Pseudocode No The paper does not contain any pseudocode or clearly labeled algorithm blocks.
Open Source Code No The paper is a conceptual discussion of technical debt in ML systems and does not present a specific methodology or tool for which source code would be released. It does not mention any open-source code for its own described work.
Open Datasets No The paper is a conceptual discussion of technical debt in ML systems and does not describe experiments that use a specific dataset for training or evaluation. It references general practices and systems but not specific public datasets used in its own research.
Dataset Splits No The paper is a conceptual work and does not conduct experiments, therefore, it does not specify any training/validation/test dataset splits.
Hardware Specification No The paper is a conceptual discussion of technical debt in ML systems and does not describe any experiments that would require specific hardware specifications.
Software Dependencies No The paper is a conceptual discussion and does not describe any specific software implementation details or experiments that would require listing software dependencies with version numbers.
Experiment Setup No The paper is a conceptual discussion and does not describe any experiments or specific model training, thus it does not provide details about experimental setup or hyperparameters.