Einsatz von Data‐Mining zur Analyse eines Polymerprozesses

Burghaus, Rolf; Leineweber, Daniel B.; Lippert, Jörg

doi:10.1002/cite.200303206

Cited by 3 publications

(3 citation statements)

References 3 publications

(1 reference statement)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The charm of these methods is not only that they can be used to identify new relationships (and this very quickly), but that it is possible to discover good descriptors [15], as well as to enlarge these techniques to further evaluations in descriptor space.…”

Section: Methods Usedmentioning

confidence: 99%

Application of Data Mining and Evolutionary Optimization in Catalyst Discovery and High‐Throughput Experimentation – Techniques, Strategies, and Software

et al. 2005

View full text Add to dashboard Cite

Data-mining and evolutionary optimization techniques are powerful tools to improve the efficiency of high-throughput experimentation (HTE) to discover new materials, drugs, or catalysts. The parameter space of screening experiments is usually high-dimensional and the parameters are possibly discrete. The response surface of the screened systems can be very rugged, characterized by smooth planes as well as steep and narrow ascents of abundant sub-optima. These conditions make exclusive use of classical statistical design and data analysis inappropriate. Evolutionary strategies, neural networks, and data mining may be an efficient alternative. Using two examples, we show the practical benefit of design strategies which combine different techniques. The selection of the methods depends on the nature of the respective HTE problem. An optimal design strategy makes HTE more efficient, and reduces research costs and time to market. Furthermore, the early application of design strategy enables reliable statements about the feasibility of the research project.

show abstract

Section: Methods Usedmentioning

confidence: 99%

Application of Data Mining and Evolutionary Optimization in Catalyst Discovery and High‐Throughput Experimentation – Techniques, Strategies, and Software

et al. 2005

View full text Add to dashboard Cite

show abstract

“…Discussion is still whether they are to be called intelligent or not; depending of course on how huge the creativity criteria are emphasized [69,70,[72][73][74]. Nevertheless, in chemical engineering AI is mostly dedicated to machine learning and further-on deep learning i.e., in majority to neural networks and data mining tools [75][76][77]. In pharmaceutical industry, expectations rank from big data mining analysis in drug development up to neuronal networks trained by operational data and used to predict maintenance, life cycles or operational parameter ranges [78][79][80][81].…”

mentioning

confidence: 99%

Accelerating Biologics Manufacturing by Modeling or: Is Approval under the QbD and PAT Approaches Demanded by Authorities Acceptable without a Digital-Twin?

et al. 2019

View full text Add to dashboard Cite

Innovative biologics, including cell therapeutics, virus-like particles, exosomes,recombinant proteins, and peptides, seem likely to substitute monoclonal antibodies as the maintherapeutic entities in manufacturing over the next decades. This molecular variety causes agrowing need for a general change of methods as well as mindset in the process development stage,as there are no platform processes available such as those for monoclonal antibodies. Moreover,market competitiveness demands hyper-intensified processes, including accelerated decisionstoward batch or continuous operation of dedicated modular plant concepts. This indicates gaps inprocess comprehension, when operation windows need to be run at the edges of optimization. Inthis editorial, the authors review and assess potential methods and begin discussing possiblesolutions throughout the workflow, from process development through piloting to manufacturingoperation from their point of view and experience. Especially, the state-of-the-art for modeling inred biotechnology is assessed, clarifying differences and applications of statistical, rigorousphysical-chemical based models as well as cost modeling. “Digital-twins” are described and effortsvs. benefits for new applications exemplified, including the regulation-demanded QbD (quality bydesign) and PAT (process analytical technology) approaches towards digitalization or industry 4.0based on advanced process control strategies. Finally, an analysis of the obstacles and possiblesolutions for any successful and efficient industrialization of innovative methods from processdevelopment, through piloting to manufacturing, results in some recommendations. A centralquestion therefore requires attention: Considering that QbD and PAT have been required byauthorities since 2004, can any biologic manufacturing process be approved by the regulatoryagencies without being modeled by a “digital-twin” as part of the filing documentation?

show abstract

“…Da es sich bei der Identifizierung von Mustern im Data Mining letztlich immer um ein kombinatorisches Optimierungsproblem von eventuell NP-Vollständigkeit handelt, kann selbst die prinzipielle Existenz von generisch tauglichen Data Mining Analyseverfahren mit Recht angezweifelt werden. Es gilt daher, spezifisch auf den Anwendungsfall zugeschnittene Analyseverfahren und -workflows zu entwickeln, die in einer möglichst großen Klasse von Anwendungsproblemen zu brauchbaren Ergebnissen führen.Hierzu wurden in den letzten Jahren erhebliche Anstrengungen unternommen, die zur Erschließung wirtschaftlich relevanter Anwendungsbereiche bei Polymerprozessen und Bioprozessen führten[2][3][4].Die ,,klassischen" Data Mining Analyseverfahren mussten dabei auf die spezifischen Bedingungen der Prozessanalyse angepasst werden:•vergleichsweise kleine Datensätze, • korrelierte, statistisch schlecht verteilte Prozessdaten, • lückenhafte Daten, • Gedächtniseffekte mit erheblicher Zeitverzögerung. Um das Potential der Data Mining Verfahren unter solchen Randbedingungen nutzen zu können, müssen spezifische Analysetechnologien entwickelt werden: • statistische Validierung von Data Mining Regelsystemen, • Stabilitäts-und Redundanzanalysen für Data Mining Regeln, • Datenvorverarbeitung, • Analyseverfahren von Zeitreihen bei Batchversuchen, • explizite Integration von Prozess-Know How in den Analyseprozess.…”

unclassified

Data Mining mit Prozessdaten (Data Mining with Process Data)

Schuppert¹,

Perne²

2005

At - Automatisierungstechnik

View full text Add to dashboard Cite

Durch die zunehmende Komplexität der Produktionsprozesse in innovativen Anwendungsfeldern wird die Prozessanalyse und -Modellierung vor neue Herausforderungen gestellt. Typische Problemfelder sind reaktive Feststoffprozesse, komplexe Polymerisationsprozesse und insbesondere biotechnologische Prozesse. Zur Analyse solcher extrem komplexer Systeme sind spezialisierte Verfahren verfügbar, die unter dem Sammelbegriff „Data Mining“ subsummiert werden. Sie erlauben eine zuverlässige Extraktion auch komplexer Muster mit hohem Informationsgehalt aus unstrukturierten Daten. Die für die Prozessanalyse relevanten softwaretechnischen und methodischen Aspekte des Data Mining werden diskutiert und entsprechende Anwendungsworkflows exemplarisch am Beispiel eines biotechnischen Produktionsprozesses sowie eines Polymerprozesses demonstriert.

show abstract

Einsatz von Data‐Mining zur Analyse eines Polymerprozesses

Abstract: No abstracts.

Cited by 3 publications

References 3 publications

Application of Data Mining and Evolutionary Optimization in Catalyst Discovery and High‐Throughput Experimentation – Techniques, Strategies, and Software

Application of Data Mining and Evolutionary Optimization in Catalyst Discovery and High‐Throughput Experimentation – Techniques, Strategies, and Software

Accelerating Biologics Manufacturing by Modeling or: Is Approval under the QbD and PAT Approaches Demanded by Authorities Acceptable without a Digital-Twin?

Data Mining mit Prozessdaten (Data Mining with Process Data)

Contact Info

Product

Resources

About