Multi-source transfer learning of time series in cyclical manufacturing

Zellinger, Werner; Grubinger, Thomas; Zwick, Michael; Lughofer, Edwin; Schöner, Holger; Natschläger, Thomas; Saminger‐Platz, Susanne

doi:10.1007/s10845-019-01499-4

Cited by 21 publications

(9 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In contrast to traditional ML approaches, TL is effective even in instances when the domains, tasks, and distributions used in training and testing are different (Lu et al 2015). To date, TL methods have been applied to many real-world scenarios, including time series forecasting and classification (Zellinger et al 2020), natural language processing (Zeng et al 2019), sentiment analysis and image recognition (Lu et al 2015;Weiss et al 2016;Flynn and Giannetti 2021). In particular, TL has gained popularity in solving image recognition problems due to advances in the field of computer vision and the availability of pre-trained models that have been trained on large image dataset such as ImageNet (Deng et al 2009).…”

Section: Transfer Learning Backgroundmentioning

confidence: 99%

“…TL has also been applied to other areas, including production modelling of time series in cyclical manufacturing and production planning. For example, a multi-source transfer learning method was proposed in Zellinger et al (2020), for modelling time series signals from sensors having different distributions. In another study, which applied TL to the field of production planning, Huang et al (2019) propose a two-stage transfer learning-based prediction method using both historical production data and real-time order data to improve accuracy and generalization performance when there are insufficient data.…”

Section: Tl In Manufacturingmentioning

confidence: 99%

“…The application of TL approaches to the manufacturing domain is an emergingbut relatively new-research area. So far, the vast majority of TL studies in the manufacturing domain address specific challenges in fault diagnosis and prognostics (Xu et al 2019;Cao et al 2020;Zhang et al 2019;Sun et al 2019;Zellinger et al 2020). In general, TL was used to: (i) leverage advances in image recognition (through reuse of pre-trained CNN models); (ii) transfer knowledge between simulated and real environments; (iii) domain adaptation, taking into account changes in data distribution for different fault conditions.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Towards scalable and reusable predictive models for cyber twins in manufacturing systems

Giannetti

Essien

2021

J Intell Manuf

View full text Add to dashboard Cite

Smart factories are intelligent, fully-connected and flexible systems that can continuously monitor and analyse data streams from interconnected systems to make decisions and dynamically adapt to new circumstances. The implementation of smart factories represents a leap forward compared to traditional automation. It is underpinned by the deployment of cyberphysical systems that, through the application of Artificial Intelligence, integrate predictive capabilities and foster rapid decision-making. Deep Learning (DL) is a key enabler for the development of smart factories. However, the implementation of DL in smart factories is hindered by its reliance on large amounts of data and extreme computational demand. To address this challenge, Transfer Learning (TL) has been proposed to promote the efficient training of models by enabling the reuse of previously trained models. In this paper, by means of a specific example in aluminium can manufacturing, an empirical study is presented, which demonstrates the potential of TL to achieve fast deployment of scalable and reusable predictive models for Cyber Manufacturing Systems. Through extensive experiments, the value of TL is demonstrated to achieve better generalisation and model performance, especially with limited datasets. This research provides a pragmatic approach towards predictive model building for cyber twins, paving the way towards the realisation of smart factories.

show abstract

Section: Transfer Learning Backgroundmentioning

confidence: 99%

Section: Tl In Manufacturingmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Towards scalable and reusable predictive models for cyber twins in manufacturing systems

Giannetti

Essien

2021

J Intell Manuf

View full text Add to dashboard Cite

show abstract

“…In this setting, various real-world applications arise, e.g. Wi-Fi localization [5], sentiment analysis [6], aerospace systems design [7], motion detection [8], and cyclical manufacturing [9]. In this work, we focus on the transfer regression problem.…”

Section: Introductionmentioning

confidence: 99%

Easy-But-Effective Domain Sub-Similarity Learning for Transfer Regression

Wei

Sagarna

et al. 2022

IEEE Trans. Knowl. Data Eng.

View full text Add to dashboard Cite

Transfer covariance function, which can model domain similarity and adaptively control the knowledge transfer across domains, is widely used in transfer learning. In this paper, we concentrate on Gaussian process (GP) models using a transfer covariance function for regression problems in a black-box learning scenario. Precisely, we investigate a family of rather general transfer covariance functions, T * , that can model the heterogeneous sub-similarities of domains through multiple kernel learning. A necessary and sufficient condition to obtain valid GPs using T * (GP T * ) for any data is given. This condition becomes specially handy for practical applications as (i) it enables semantic interpretations of the sub-similarities and (ii) it can readily be used for model learning. In particular, we propose a computationally inexpensive model learning rule that can explicitly capture different sub-similarities of domains. We propose two instantiations of GP T * , one with a set of predefined constant base kernels and one with a set of learnable parametric base kernels. Extensive experiments on 36 synthetic transfer tasks and 12 real-world transfer tasks demonstrate the effectiveness of GP T * on the sub-similarity capture and the transfer performance.

show abstract

“…Figure 1 for an illustration. Moment-based algorithms perform particularly well in many practical tasks [21,4,53,65,67,66,30,34,68,42,45,28,63,64,46,44]. Second, by the current scientific discussion about the choice of an appropriate distance function for domain adaptation [8,15,36,37,69,24].…”

mentioning

confidence: 99%

On generalization in moment-based domain adaptation

Zellinger

Moser

Saminger‐Platz

2020

Ann Math Artif Intell

Self Cite

View full text Add to dashboard Cite

Domain adaptation algorithms are designed to minimize the misclassification risk of a discriminative model for a target domain with little training data by adapting a model from a source domain with a large amount of training data. Standard approaches measure the adaptation discrepancy based on distance measures between the empirical probability distributions in the source and target domain. In this setting, we address the problem of deriving learning bounds under practice-oriented general conditions on the underlying probability distributions. As a result, we obtain learning bounds for domain adaptation based on finitely many moments and smoothness conditions. MotivationDomain adaptation problems are encountered in everyday life of engineering machine learning applications whenever there is a discrepancy between assumptions on the learning and application setting. For example, most theoretical and practical results in statistical learning are based on the assumption that the training and test sample are drawn from the same distribution. As outlined in [11,7,43,10], however, this assumption may be violated in typical applications such as natural language processing [12, 27] and computer vision [24, 65, 67].

show abstract

Multi-source transfer learning of time series in cyclical manufacturing

Cited by 21 publications

References 38 publications

Towards scalable and reusable predictive models for cyber twins in manufacturing systems

Towards scalable and reusable predictive models for cyber twins in manufacturing systems

Easy-But-Effective Domain Sub-Similarity Learning for Transfer Regression

On generalization in moment-based domain adaptation

Contact Info

Product

Resources

About