Style Editor Docs
     
  • Home
  • Products & Services
    • DataHowLab
    • Professional Services
    • Innovation Hub
  • Resources
    • Technology
    • Publications
    • Bioprocess Blog
  • Education & Events
    • Schedule events & trainings
    • Trainings
    • Upcoming Events
  • About
    • DataHow
    • Team
    • Careers
    • Contact
   info@datahow.ch
  • Home
  • Products & Services
    • DataHowLab
    • Professional Services
    • Innovation Hub
  • Resources
    • Technology
    • Publications
    • Bioprocess Blog
  • Education & Events
    • Schedule events & trainings
    • Trainings
    • Upcoming Events
  • About
    • DataHow
    • Team
    • Careers
    • Contact
    Publications  ·  06. July 2021

    Hybrid Models Based on Machine Learning and an Increasing Degree of Process Knowledge: Application to Capture Chromatographic Step

    Full title: 

    Hybrid Models Based on Machine Learning and an Increasing Degree of Process Knowledge: Application to Capture Chromatographic Step

     

    Abstract:

    In process engineering, two paradigms of modeling approaches exist: the mechanistic and the data-driven approaches with the former being completely based on knowledge while the latter completely based on data. In our previous work, we highlighted the advantages of using hybrid models that explores the synergy between mechanistic and data-driven models. Here we introduce the concept of developing a series of hybrid models constituted by a progressively increasing extent of process knowledge. Thus, aligning the models on the “degrees of hybridization” axis with data-driven model being 0% hybridized and mechanistic model being 100% hybridized. In this work, the proposed concept is demonstrated for the application of a chromatographic capture step where the models are evaluated based on (i) prediction accuracy, (ii) extrapolation ability, (iii) providing process understanding, and (iv) practical application. We show the limitations of both model variant extremes. On one hand, the performance of the mechanistic model is compromised due to an excessive imposition of knowledge, thus affecting its predictive capabilities and efficiency in practical utility. On the other hand, the data-driven model inherently is not suitable for application such as multicolumn chromatography or to gain process understanding. In contrast, a series of hybrid models could be developed with better and versatile performance in term of prediction, extrapolation, process understanding, and practical utility. We show that for general process applications the different hybrid model variants and their ensembles have comparable performance. We illustrate the criteria for selection of a particular hybrid model variant based on different considerations such as complexity of training or model development, acquired understanding, and data requirement.

    Full Publication
    tagPlaceholderTags:

    Write a comment

    Comments: 0
    draggable-logo

    Hagenholzstrasse.111 / 8050 Zurich/ info@datahow.ch 


    Schedule a Demo

    Edit here your navigation button
    is-switcher admin-only
    is-switcher admin-only
    is-switcher admin-only

     

    Main colors
       bg-primary
       bg-primary-light
       bg-primary-dark
       bg-secondary
       bg-secondary-dark
    Template sections
       body
       top-header
       header
       content
    Footer Styles
       background
       text color
       link color
       horizontal line
    Buttons
       style 1
       style 2
       style 3
    Other elements
      social icons
      navigation color
      subnav background
    Mobile navigation
       background color
       navigation color
    Template configurations
    g-font
    navigation styles
    size-17 weight-400
    content styles
    form-white
    footer styles
    o-form color-white
    Typography
    Heading H1
    weight-600
    Heading H2
    weight-600
    Heading H3
    weight-600
    Buttons
    weight-600 is-uppercase
    Animations
    wow animated fadeInUp

    Note:
    All changes made here will be applied to your entire website.

    About | Privacy Policy | Cookie Policy | Sitemap
    © Copyright 2022 - DataHow - All rights reserved
    Log out | Edit
    • Scroll to top