| 1. |
|
|
| 2. |
- Bayram, Firas, et al.
(author)
-
A Drift Handling Approach for Self-Adaptive ML Software in Scalable Industrial Processes
- 2022
-
In: Proceedings of the 37th IEEE/ACM International Conference on Automated Software Engineering. - New York, NY, USA : Association for Computing Machinery (ACM). - 9781450394758 ; , s. 1-5
-
Conference paper (peer-reviewed)abstract
- Most industrial processes in real-world manufacturing applications are characterized by the scalability property, which requires an automated strategy to self-adapt machine learning (ML) software systems to the new conditions. In this paper, we investigate an Electroslag Remelting (ESR) use case process from the Uddeholms AB steel company. The use case involves predicting the minimum pressure value for a vacuum pumping event. Taking into account the long time required to collect new records and efficiently integrate the new machines with the built ML software system. Additionally, to accommodate the changes and satisfy the non-functional requirement of the software system, namely adaptability, we propose an automated and adaptive approach based on a drift handling technique called importance weighting. The aim is to address the problem of adding a new furnace to production and enable the adaptability attribute of the ML software. The overall results demonstrate the improvements in ML software performance achieved by implementing the proposed approach over the classical non-adaptive approach.
|
|
| 3. |
- Bayram, Firas, et al.
(author)
-
DQSOps : Data Quality Scoring Operations Framework for Data-Driven Applications
- 2023
-
In: EASE '23: Proceedings of the 27<sup>th</sup> International Conference on Evaluation and Assessment in Software Engineering. - : Association for Computing Machinery (ACM). - 9798400700446 ; , s. 32-41
-
Conference paper (peer-reviewed)abstract
- Data quality assessment has become a prominent component in the successful execution of complex data-driven artificial intelligence (AI) software systems. In practice, real-world applications generate huge volumes of data at speeds. These data streams require analysis and preprocessing before being permanently stored or used in a learning task. Therefore, significant attention has been paid to the systematic management and construction of high-quality datasets. Nevertheless, managing voluminous and high-velocity data streams is usually performed manually (i.e. offline), making it an impractical strategy in production environments. To address this challenge, DataOps has emerged to achieve life-cycle automation of data processes using DevOps principles. However, determining the data quality based on a fitness scale constitutes a complex task within the framework of DataOps. This paper presents a novel Data Quality Scoring Operations (DQSOps) framework that yields a quality score for production data in DataOps workflows. The framework incorporates two scoring approaches, an ML prediction-based approach that predicts the data quality score and a standard-based approach that periodically produces the ground-truth scores based on assessing several data quality dimensions. We deploy the DQSOps framework in a real-world industrial use case. The results show that DQSOps achieves significant computational speedup rates compared to the conventional approach of data quality scoring while maintaining high prediction performance.
|
|
| 4. |
- Chatterjee, Ayan, et al.
(author)
-
Testing of machine learning models with limited samples : an industrial vacuum pumping application
- 2022
-
In: ESEC/FSE ’22-Proceedings of the 30<sup>th</sup> ACM Joint Meeting European Software Engineering Conference and Symposium on the Foundations of Software Engineering. - New York, NY, USA : Association for Computing Machinery (ACM). - 9781450394130 ; , s. 1280-1290
-
Conference paper (peer-reviewed)abstract
- There is often a scarcity of training data for machine learning (ML) classification and regression models in industrial production, especially for time-consuming or sparsely run manufacturing processes. Traditionally, a majority of the limited ground-truth data is used for training, while a handful of samples are left for testing. In that case, the number of test samples is inadequate to properly evaluate the robustness of the ML models under test (i.e., the system under test) for classification and regression. Furthermore, the output of these ML models may be inaccurate or even fail if the input data differ from the expected. This is the case for ML models used in the Electroslag Remelting (ESR) process in the refined steel industry to predict the pressure in a vacuum chamber. A vacuum pumping event that occurs once a workday generates a few hundred samples in a year of pumping for training and testing. In the absence of adequate training and test samples, this paper first presents a method to generate a fresh set of augmented samples based on vacuum pumping principles. Based on the generated augmented samples, three test scenarios and one test oracle are presented to assess the robustness of an ML model used for production on an industrial scale. Experiments are conducted with real industrial production data obtained from Uddeholms AB steel company. The evaluations indicate that Ensemble and Neural Network are the most robust when trained on augmented data using the proposed testing strategy. The evaluation also demonstrates the proposed method's effectiveness in checking and improving ML algorithms' robustness in such situations. The work improves software testing's state-of-the-art robustness testing in similar settings. Finally, the paper presents an MLOps implementation of the proposed approach for real-time ML model prediction and action on the edge node and automated continuous delivery of ML software from the cloud.
|
|
| 5. |
- Langenskiöld, Marcus, 1972, et al.
(author)
-
Deep Femoral Vein Reconstruction for Abdominal Aortic Graft Infections is Associated with Low Aneurysm Related Mortality and a High Rate of Permanent Discontinuation of Antimicrobial Treatment
- 2021
-
In: European Journal of Vascular and Endovascular Surgery. - : Elsevier. - 1078-5884 .- 1532-2165. ; 62:6, s. 927-934
-
Journal article (peer-reviewed)abstract
- Objective: Aortic prosthesis infection is a devastating complication of aortic surgery. In situ reconstruction with the neo-aorto-iliac system (NAIS) bypass technique has become increasingly used and is recommended in recent treatment guidelines. The main aim was to evaluate NAIS procedural outcomes when undertaken after previous open or endovascular aortic repair in Sweden.Methods: In this retrospective study, The National Quality Registry for Vascular Surgery (Swedvasc) was used to identify Swedish centres that offered the NAIS bypass procedure for aortic prosthesis infection between 2008 and 2018. Variables of special interest were procedural details, short and long term survival, renal and other complications, and the durtion of antimicrobial treatment.Results: Forty patients (36 males, four females [mean age 69 years], 32 open repairs, seven endovascular aortic repairs [EVAR] and one fenestrated EVAR; 21 presented with aorto-enteric fistula) operated on with NAIS bypass were reviewed. The median time from the primary aortic intervention to the NAIS bypass procedure was 32 months (range 0 – 252 months). Mean ± standard deviation operating time was 645 ± 160 minutes, mean blood loss was 6 277 ± 6 525 mL, mean length of intensive care unit stay was 5.3 ± 3.7 days, and mean length of overall hospital stay was 21.2 ± 11.4 days. Thirty-five patients (88%) had a positive microbial culture; the most commonly isolated pathogen was Candida spp. The majority of patients survived for 30 days (n = 35 [88%]), and 33 (83%) and 32 (80%) patients survived for 90 days and one year, respectively. The number of surviving patients free from antimicrobial treatment at 90 days, six months, and one year was 19 (58%), 29 (88%), and 30 (94%). After a mean long term follow up of 69.9 ± 44.7 months, 20 patients were still alive.Conclusion: The NAIS bypass procedure offered reasonable survival and functional outcomes, and was associated with a high cure rate, defined as freedom from any antimicrobial treatment.
|
|
| 6. |
- Monteux, Sylvain, et al.
(author)
-
Carbon and nitrogen cycling in Yedoma permafrost controlled by microbial functional limitations
- 2020
-
In: Nature Geoscience. - : Nature Publishing Group. - 1752-0894 .- 1752-0908. ; 13:12, s. 794-
-
Journal article (peer-reviewed)abstract
- Warming-induced microbial decomposition of organic matter in permafrost soils constitutes a climate-change feedback of uncertain magnitude. While physicochemical constraints on soil functioning are relatively well understood, the constraints attributable to microbial community composition remain unclear. Here we show that biogeochemical processes in permafrost can be impaired by missing functions in the microbial community-functional limitations-probably due to environmental filtering of the microbial community over millennia-long freezing. We inoculated Yedoma permafrost with a functionally diverse exogenous microbial community to test this mechanism by introducing potentially missing microbial functions. This initiated nitrification activity and increased CO2 production by 38% over 161 days. The changes in soil functioning were strongly associated with an altered microbial community composition, rather than with changes in soil chemistry or microbial biomass. The present permafrost microbial community composition thus constrains carbon and nitrogen biogeochemical processes, but microbial colonization, likely to occur upon permafrost thaw in situ, can alleviate such functional limitations. Accounting for functional limitations and their alleviation could strongly increase our estimate of the vulnerability of permafrost soil organic matter to decomposition and the resulting global climate feedback. Carbon dioxide emissions from permafrost thaw are substantially enhanced by relieving microbial functional limitations, according to incubation experiments on Yedoma permafrost.
|
|
| 7. |
- Stenberg, Simon, et al.
(author)
-
Genetically controlled mtDNA deletions prevent ROS damage by arresting oxidative phosphorylation
- 2022
-
In: eLife. - : eLife Sciences Publications, Ltd. - 2050-084X. ; 11
-
Journal article (peer-reviewed)abstract
- Deletion of mitochondrial DNA in eukaryotes is currently attributed to rare accidental events associated with mitochondrial replication or repair of double-strand breaks. We report the discovery that yeast cells arrest harmful intramitochondrial superoxide production by shutting down respiration through genetically controlled deletion of mitochondrial oxidative phosphorylation genes. We show that this process critically involves the antioxidant enzyme superoxide dismutase 2 and two-way mitochondrial-nuclear communication through Rtg2 and Rtg3. While mitochondrial DNA homeostasis is rapidly restored after cessation of a short-term superoxide stress, long-term stress causes maladaptive persistence of the deletion process, leading to complete annihilation of the cellular pool of intact mitochondrial genomes and irrevocable loss of respiratory ability. This shows that oxidative stress-induced mitochondrial impairment may be under strict regulatory control. If the results extend to human cells, the results may prove to be of etiological as well as therapeutic importance with regard to age-related mitochondrial impairment and disease.
|
|
| 8. |
- Stenberg, Simon, et al.
(author)
-
Genetically controlled mtDNA deletions prevent ROS damage by arresting oxidative phosphorylation.
- 2022
-
In: eLife. - 2050-084X. ; 11
-
Journal article (peer-reviewed)abstract
- Deletion of mitochondrial DNA in eukaryotes is currently attributed to rare accidental events associated with mitochondrial replication or repair of double-strand breaks. We report the discovery that yeast cells arrest harmful intramitochondrial superoxide production by shutting down respiration through genetically controlled deletion of mitochondrial oxidative phosphorylation genes. We show that this process critically involves the antioxidant enzyme superoxide dismutase 2 and two-way mitochondrial-nuclear communication through Rtg2 and Rtg3. While mitochondrial DNA homeostasis is rapidly restored after cessation of a short-term superoxide stress, long-term stress causes maladaptive persistence of the deletion process, leading to complete annihilation of the cellular pool of intact mitochondrial genomes and irrevocable loss of respiratory ability. This shows that oxidative stress-induced mitochondrial impairment may be under strict regulatory control. If the results extend to human cells, the results may prove to be of etiological as well as therapeutic importance with regard to age-related mitochondrial impairment and disease.
|
|
| 9. |
- Thompson, Lo Hallin, et al.
(author)
-
Impact of Adrenalectomy on Morbidity in Patients with Non-Functioning Adrenal Cortical Tumours, Mild Hypercortisolism and Cushing’s Syndrome as Assessed by National and Quality Registries
- 2021
-
In: World Journal of Surgery. - : Springer Science and Business Media LLC. - 0364-2313 .- 1432-2323. ; 45:10, s. 3099-3107
-
Journal article (peer-reviewed)abstract
- Background: The impact of adrenalectomy on morbidity in patients with mild hypercortisolism and non-functioning adrenocortical adenoma is unclear. The present study evaluated morbidity before and after adrenalectomy in patients with benign adrenocortical tumour with Cushing´s syndrome (CS), autonomous cortisol secretion (ACS) and non-functioning adrenocortical adenoma as assessed by national and quality registries. Methods: Patients registered in the Scandinavian Quality Register for Thyroid, Parathyroid and Adrenal Surgery (SQRTPA) 2009–2017 with CS, ACS or non-functioning adrenocortical adenoma, were included in this retrospective study and analysed with age- and sex-matched controls, 1:3. Morbidity associated with CS was assessed pre- and postoperatively by analysing data from the Swedish National Patient Register and the Swedish Prescribed Drug Register. Results: Some 271 patients were included, CS (127), ACS (45) and non-functioning adrenocortical adenoma (99), with 813 matched controls. The frequency of hypertension was almost 50% in all tumour groups. Antihypertensive medication preoperatively was more frequent in all tumour groups compared with controls. No preoperative differences in medication were detected between patients with CS and ACS. A decrease in the use of hypertensive drugs was noticed annually for all patient groups after adrenalectomy. Conclusions: Hypertension is common in patients with benign adrenocortical tumours regardless of cortisol hypersecretion. The use of antihypertensive drugs in patients with CS, ACS and non-functioning adrenocortical adenoma was reduced after adrenalectomy. These findings highlight the need for a randomized controlled trial to investigate the impact of adrenalectomy on morbidity in patients with mild hypercortisolism.
|
|
