Filters








11,427 Hits in 4.1 sec

Hidden Technical Debt in Machine Learning Systems

D Sculley, Gary Holt, Daniel Golovin, Eugene Davydov, Todd Phillips, Dietmar Ebner, Vinay Chaudhary, Michael Young, Jean-François Crespo, Dan Dennison
unpublished
Using the software engineering framework of technical debt, we find it is common to incur massive ongoing maintenance costs in real-world ML systems.  ...  Machine learning offers a fantastically powerful toolkit for building useful complex prediction systems quickly. This paper argues it is dangerous to think of these quick wins as coming for free.  ...  Acknowledgments This paper owes much to the important lessons learned day to day in a culture that values both innovative ML research and strong engineering practice.  ... 
fatcat:7sbcfequrba6hjxyqlsauigoxu

Quality Assurance Challenges for Machine Learning Software Applications During Software Development Life Cycle Phases [article]

Md Abdullah Al Alamin, Gias Uddin
2021 arXiv   pre-print
In the past decades, the revolutionary advances of Machine Learning (ML) have shown a rapid adoption of ML models into software systems of diverse types.  ...  Such Machine Learning Software Applications (MLSAs) are gaining importance in our daily lives. As such, the Quality Assurance (QA) of MLSAs is of paramount importance.  ...  The latest trend is the widespread interest in the adoption of ML (Machine Learning) capabilities into large scale Machine Learning software applications (MLSAs) [1] .  ... 
arXiv:2105.01195v2 fatcat:ksvuixfvxjeddpky4ifmnnh5re

Technology Readiness Levels for AI ML [article]

Alexander Lavin, Gregory Renard
2020 arXiv   pre-print
Drawing on experience in both spacecraft engineering and AI/ML (from research through product), we propose a proven systems engineering approach for machine learning development and deployment.  ...  The development and deployment of machine learning systems can be executed easily with modern tools, but the process is typically rushed and means-to-an-end.  ...  We've introduced TRL4ML, a proven systems engineering process for machine learning.  ... 
arXiv:2006.12497v3 fatcat:nn2co36ypbewvlpfrqu7vuc3gq

Design of a Financial Decision Support System based on Artificial Neural Networks for Stock Price Prediction

Sandeep Patalay
2019 JOURNAL OF MECHANICS OF CONTINUA AND MATHEMATICAL SCIENCES  
Machine learning is widely being used in the financial domain including prediction of stock prices.  ...  Machine learning focuses on the development of computer programs that can access data and use it to learn for themselves.  ...  parameters which will enable the system to comprehensively predict stock prices.The authors are working on enhancing the machine learning models and fine tuning the financial parameters to more accurately  ... 
doi:10.26782/jmcms.2019.10.00060 fatcat:h5u2jom7ajbdhh3st3h7shvsg4

A Debt-Aware Learning Approach for Resource Adaptations in Cloud Elasticity Management [article]

Carlos Mera-Gómez, Francisco Ramírez, Rami Bahsoon, Rajkumar Buyya
2017 arXiv   pre-print
To address this limitation, we propose a technical debt-aware learning approach for autonomous elasticity management based on a reinforcement learning of elasticity debts in resource provisioning; the  ...  The evaluation shows that a reinforcement learning of technical debts in elasticity obtains a higher utility for a cloud customer, while conforming expected levels of performance.  ...  We instantiated from the reference system model in figure 2 a scenario with a single debt-aware learning agent. A.  ... 
arXiv:1702.07431v1 fatcat:d3fbejh2qzf3hmm2buwdc2shdm

Cognitive Computing Architectures for Machine (Deep) Learning at Scale

Samir Mittal
2017 Proceedings (MDPI)  
The paper reviews existing models for organizing information for machine learning systems in heterogeneous computing environments.  ...  We extend these concepts to the design of Cognitive Distributed Learning Systems to resolve critical bottlenecks in real-time machine learning applications such as Predictive Analytics and Recommender  ...  Figure 4 . 4 Computational requirements for scale-out learning (from [5] ). Figure 5 . 5 The hidden technical debt in machine learning systems (adapted from [6] ).  ... 
doi:10.3390/is4si-2017-04025 fatcat:3ieob5bxwnd5hgjb7mytoxyszi

Making Contextual Decisions with Low Technical Debt [article]

Alekh Agarwal, Sarah Bird, Markus Cozowicz, Luong Hoang, John Langford, Stephen Lee, Jiaji Li, Dan Melamed, Gal Oshri, Oswaldo Ribas, Siddhartha Sen, Alex Slivkins
2017 arXiv   pre-print
Reinforcement-based learning algorithms such as contextual bandits can be very effective in these settings, but applying them in practice is fraught with technical debt, and no general system exists that  ...  The service makes real-time decisions and learns continuously and scalably, while significantly lowering technical debt.  ...  Systems for ML and experimentation We previously discussed these systems with regards to technical debt ( §2). A more general discussion follows. A/B testing.  ... 
arXiv:1606.03966v2 fatcat:6aafbkrqaregbhykmw525o4lqq

What do you mean? Research in the Age of Machines

Arthur "A.J." Boston
2019 College & research libraries news  
was an undeniable bop of its era in which Justin Bieber explores the ambiguities of romantic communication.  ...  (I pinky promise this will soon make sense for scholarly communication librarians interested in artificial intelligence [AI].)  ...  Data mining and machine-learning are a great boon for political campaigns and corporate marketing wings that thrive on the ability to uncover hidden connections in consumer behavior, in order to influence  ... 
doi:10.5860/crln.80.10.565 fatcat:2l3lh3yaavck3fzid6uzzrbluq

Share Price Prediction of Aerospace Relevant Companies with Recurrent Neural Networks based on PCA [article]

Linyu Zheng, Hongmei He
2020 arXiv   pre-print
The developed model could be an intelligent agent in an automatic stock prediction system, with which, the financial industry could make a prompt decision for their economic strategies and business activities  ...  The capital market plays a vital role in marketing operations for aerospace industry.  ...  Moreover, various machine-learning techniques have been widely used in the finance industry.  ... 
arXiv:2008.11788v1 fatcat:p4euefg5e5gv5dbriy4afo635a

A Framework for Conditional Statement Technical Debt Identification and Description [article]

Abdulaziz Alhefdhi
2022 arXiv   pre-print
However, there are many cases where technical debt instances are not explicitly acknowledged but deeply hidden in the code.  ...  While our approach is applicable in principle to any type of code fragments, we focus in this study on technical debt hidden in conditional statements, one of the most TD-carrying parts of code.  ...  This investigation ensured that the SATD comments represent technical debt in source code.  ... 
arXiv:2012.12466v2 fatcat:vykl4yb6kfaefoiikpo3mhh3qy

Prevalence, Contents and Automatic Detection of KL-SATD [article]

Leevi Rantala, Mika Mäntylä, David Lo
2020 arXiv   pre-print
When developers use different keywords such as TODO and FIXME in source code comments to describe self-admitted technical debt (SATD), we refer it as Keyword-Labeled SATD (KL-SATD).  ...  Our machine learning classifier using logistic Lasso regression has good performance in detecting KL-SATD comments (AUC-ROC 0.88).  ...  Selfadmitted technical debt (SATD) refers to a specific type of code debt, where the developer acknowledges admitting code debt into the system [2] .  ... 
arXiv:2008.05159v1 fatcat:mh5kzlopsbc7zajschcn25zl5e

Data Smells in Public Datasets [article]

Arumoy Shome and Luis Cruz and Arie van Deursen
2022 arXiv   pre-print
Analogous to code smells, we introduce a novel catalogue of data smells that can be used to indicate early signs of problems or technical debt in machine learning systems.  ...  The adoption of Artificial Intelligence (AI) in high-stakes domains such as healthcare, wildlife preservation, autonomous driving and criminal justice system calls for a data-centric approach to AI.  ...  Due to their highly tangled and experimental nature, machine learning systems are prone to rapid accumulation of technical debt [4, 5, 8, 46] .  ... 
arXiv:2203.08007v2 fatcat:5mhual47krfflg5bjsu3wefxle

THE EVOLUTION OF KDD: TOWARDS DOMAIN-DRIVEN DATA MINING

LONGBING CAO, CHENGQI ZHANG
2007 International journal of pattern recognition and artificial intelligence  
Key components of domain-driven data mining are constrained context, integrating domain intelligence, human-machine cooperation, in-depth mining, actionability enhancement, and iterative refinement process  ...  Its typical task is to let data tell a story disclosing hidden information regarding a business issue.  ...  Data used in this paper is from Australian Centrelink and Capital Market CRC, AC3 and SIRCA.  ... 
doi:10.1142/s0218001407005612 fatcat:6ln5hcaxefaftp44kjdywbpjy4

Debt-Prone Bugs: Technical Debt in Software Maintenance [article]

Jifeng Xuan, Yan Hu, He Jiang
2017 arXiv   pre-print
In this article, we propose the concept of debt-prone bugs to model the technical debt in software maintenance.  ...  Such confliction leads to a trade-off between software quality and release schedule, which is known as the technical debt metaphor.  ...  [1] summarize the state of art in managing technical debt in software systems and present the potential future work. In bug tracking systems, Storey et al. [9] , Wang et al. [10] , and Guo et al.  ... 
arXiv:1704.04766v1 fatcat:ezkgllmpnvc7xidncd22g5ypc4

Debt-Prone Bugs: Technical Debt in Software Maintenance

Jifeng Xuan, Yan Hu, He Jiang
2012 International Journal of Advancements in Computing Technology  
In this article, we propose the concept of debt-prone bugs to model the technical debt in software maintenance.  ...  Such confliction leads to a trade-off between software quality and release schedule, which is known as the technical debt metaphor.  ...  [1] summarize the state of art in managing technical debt in software systems and present the potential future work. In bug tracking systems, Storey et al. [9] , Wang et al. [10] , and Guo et al.  ... 
doi:10.4156/ijact.vol4.issue19.54 fatcat:gtp3rbrkincdlp36chrq7iurxm
« Previous Showing results 1 — 15 out of 11,427 results