Machine Learning Operations (MLOps): Overview, Definition, and Architecture

Kreuzberger, Dominik; Kuehl, Niklas; Hirschl, Sebastian

doi:10.1109/access.2023.3262138

Cited by 150 publications

(49 citation statements)

References 60 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Finally, from a scalability and operationalization standpoint, considering the deliberate modularity of AIPCO and its absence from pathologist annotations, we envision leveraging the full potential of machine learning operations (MLOPs) [49] to create an automated machine learning solution capable of generating candidate algorithms. On the one hand, the adoption of MLOPs would lead to the creation of an H&E-based IO-predictive algorithm factory.…”

Section: Discussionmentioning

confidence: 99%

Immunotherapy Efficacy Prediction in Cancer: An Artificial Intelligence Approach with Unannotated H&E Whole-Slide Images

Conde,

Qaiser,

et al. 2024

Preprint

View full text Add to dashboard Cite

Developing a solution to predict clinical outcomes for immunotherapy that is accurate, scalable, affordable, clinically meaningful, and globally accessible is an unmet medical need. Precise prediction of patient response to immunotherapy from pretreatment biopsy images will enable the delivery of immuno-oncology drugs to suitable patients and reduce the risk of administering unnecessary toxicity to patients who do not benefit from it. We propose an AI-based framework to produce stratifying algorithms that only need routinely obtained unannotated hematoxylin and eosin (H&E)-stained whole slide images. This design choice eliminates the need for pathologist annotations, ensuring affordability and scalability. Our solution, developed with data from 418 durvalumab patients, was validated both for head and neck squamous cell carcinoma and non-small cell lung cancer with data from 283 durvalumab patients, demonstrating its versatility and ease of adaptation for different indications. The results obtained using test data from clinical trials, different from training data, exhibit clinically meaningful improvement between those classified as positive and negative. For median overall survival (OS), the enhancement is in the range [55.9%, 198%] and [0.49, 0.70] for the hazard ratio for OS. For median progression-free survival (PFS), the improvement ranges within [39%, 195%], while the hazard ratio is within [0.46, 0.86] for PFS. Our solution complements the current biomarker, programmed death lig– and 1, for immunotherapy therapy, presenting an opportunity to develop more accurate solutions. In addition, as the algorithm was developed in a hypothesis-free approach, the analysis of the converged solution may enhance expert understanding of the pathomechanisms driving the response to immunotherapy. Its scalability and accuracy, combined with the AI-based engineering nature of the solution, bring the opportunity of being globally deployed using the cloud. Our technique has the potential to unlock opportunities not available for patients before by enabling the generation of efficient and affordable diagnoses for immunotherapy.

show abstract

Section: Discussionmentioning

confidence: 99%

Immunotherapy Efficacy Prediction in Cancer: An Artificial Intelligence Approach with Unannotated H&E Whole-Slide Images

Conde,

Qaiser,

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…This approach is applicable to any type of machine learning model, with a focus on neural networks [3] . Moreschini et al propose a graphical representation for MLOps, called MLOps, which combines the simplicity of DevOps with circular steps for ML incorporation, creating a self-maintained ML-based development subsystem [4] . Finally, Cankar et al address the security concerns in DevOps by proposing IaC Scan Runner and LOMOS, tools that provide static analysis and runtime anomaly detection for Infrastructure as Code (IaC) [5] .…”

Section: Literature Reviewmentioning

confidence: 99%

Unlocking the Potential of AI/ML in DevSecOps: Effective Strategies and Optimal Practices

Camacho

2024

JAIGS

View full text Add to dashboard Cite

In the dynamic realm of technology, the fusion of Artificial Intelligence (AI) and Machine Learning (ML) with DevSecOps practices stands out as a pivotal catalyst for bolstering security, efficiency, and innovation in software development and deployment processes. This document explores effective strategies and optimal practices for maximizing the capabilities of AI/ML within the DevSecOps framework. Commencing with an overview of DevSecOps principles and the integral role of AI/ML, the document delves into specific tactics such as automated threat detection, predictive analytics for vulnerability management, and intelligent automation for continuous integration and deployment. Additionally, it addresses prominent challenges and considerations associated with the integration of AI/ML in DevSecOps, including data privacy, algorithm transparency, and ethical implications. Through illuminating case studies and real-world illustrations, the document showcases how organizations can leverage AI/ML technologies to streamline their DevSecOps pipelines, mitigate security risks, and cultivate a culture of ongoing enhancement. By embracing these strategies and adhering to best practices, organizations can harness the full potential of AI/ML to propel innovation, fortify resilience, and enhance agility in their DevSecOps endeavors.

show abstract

“…A thematic area close to AutoML concerns the ML Operations (MLOps) prac MLOps constitute a set of tools and mechanisms able to enhance the collaboratio tween data scientists and IT professionals in applying and maintaining ML models [2 In turn, MLOps practices are in the position to assist the risk management associated deploying ML models in production by providing traceability, monitoring, and te capabilities [28,30,31]. In general, two major issues are considered in designing M procedures [28,30,32]. The first one is related to compatibility problems arising from fact that MLOps practices integrate a wide range of heterogeneous tools, technolo and software environments.…”

Section: Introductionmentioning

confidence: 99%

“…In turn, MLOps practices are in the position to assist the risk management associated with deploying ML models in production by providing traceability, monitoring, and testing capabilities [28,30,31]. In general, two major issues are considered in designing MLOps procedures [28,30,32]. The first one is related to compatibility problems arising from the fact that MLOps practices integrate a wide range of heterogeneous tools, technologies, and software environments.…”

Section: Introductionmentioning

confidence: 99%

“…The former optimizes the model's development and deployment in terms of continuous integration and continuous delivery (CI/CD) practices. The latter enhances the production of highly effective and reliable ML models [28,33]. CI/CD mechanisms are very essential in MLOps' development because they establish real-time model testing and deployment [34].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Structure Learning and Hyperparameter Optimization Using an Automated Machine Learning (AutoML) Pipeline

et al. 2023

View full text Add to dashboard Cite

In this paper, we built an automated machine learning (AutoML) pipeline for structure-based learning and hyperparameter optimization purposes. The pipeline consists of three main automated stages. The first carries out the collection and preprocessing of the dataset from the Kaggle database through the Kaggle API. The second utilizes the Keras-Bayesian optimization tuning library to perform hyperparameter optimization. The third focuses on the training process of the machine learning (ML) model using the hyperparameter values estimated in the previous stage, and its evaluation is performed on the testing data by implementing the Neptune AI. The main technologies used to develop a stable and reusable machine learning pipeline are the popular Git version control system, the Google cloud virtual machine, the Jenkins server, the Docker containerization technology, and the Ngrok reverse proxy tool. The latter can securely publish the local Jenkins address as public through the internet. As such, some parts of the proposed pipeline are taken from the thematic area of machine learning operations (MLOps), resulting in a hybrid software scheme. The machine learning model was used to evaluate the pipeline, which is a multilayer perceptron (MLP) that combines typical dense, as well as polynomial, layers. The simulation results show that the proposed pipeline exhibits a reliable and accurate performance while managing to boost the network’s performance in classification tasks.

show abstract

Machine Learning Operations (MLOps): Overview, Definition, and Architecture

Cited by 150 publications

References 60 publications

Immunotherapy Efficacy Prediction in Cancer: An Artificial Intelligence Approach with Unannotated H&E Whole-Slide Images

Immunotherapy Efficacy Prediction in Cancer: An Artificial Intelligence Approach with Unannotated H&E Whole-Slide Images

Unlocking the Potential of AI/ML in DevSecOps: Effective Strategies and Optimal Practices

Structure Learning and Hyperparameter Optimization Using an Automated Machine Learning (AutoML) Pipeline

Contact Info

Product

Resources

About