Data-centric Engineering: integrating simulation, machine learning and statistics. Challenges and opportunities

Pan, Indranil; Mason, Lachlan; Matar, Omar

doi:10.1016/j.ces.2021.117271

Cited by 32 publications

(15 citation statements)

References 54 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Due to our focus on synthetic data, we have not included any data-centric methods concerned with data labeling. However, we note that significant work in the ML community has focused on the effects of data labeling and developed a set of data-centric approaches for systematically relabeling or removing mislabeled data to improve overall system performance [47,[65][66][67]. An interesting problem for future studies would be to consider the application of these label-focused approaches to experimental QIS systems.…”

Section: Discussionmentioning

confidence: 99%

Data-centric machine learning in quantum information science

Lohani

Lukens

Glasser

et al. 2022

Mach. Learn.: Sci. Technol.

View full text Add to dashboard Cite

We propose a series of data-centric heuristics for improving the performance of machine learning systems when applied to problems in quantum information science. In particular, we consider how systematic engineering of training sets can significantly enhance the accuracy of pre-trained neural networks used for quantum state reconstruction without altering the underlying architecture. We find that it is not always optimal to engineer training sets to exactly match the expected distribution of a target scenario, and instead, performance can be further improved by biasing the training set to be slightly more mixed than the target. This is due to the heterogeneity in the number of free variables required to describe states of different purity, and as a result, overall accuracy of the network improves when training sets of a fixed size focus on states with the least constrained free variables. For further clarity, we also include a ``toy model'' demonstration of how spurious correlations can inadvertently enter synthetic data sets used for training, how the performance of systems trained with these correlations can degrade dramatically, and how the inclusion of even relatively few counterexamples can effectively remedy such problems.

show abstract

Section: Discussionmentioning

confidence: 99%

Data-centric machine learning in quantum information science

Lohani

Lukens

Glasser

et al. 2022

Mach. Learn.: Sci. Technol.

View full text Add to dashboard Cite

show abstract

“…As ML developers in areas outside medicine are increasingly noticing, the next step in the development of clinically deployable AI will require a more data-centric approach than the hitherto dominating model-centric approach, where lack of data quality is sought to be compensated for by increased volumes of data. [38,39] The capability of modern ML models to process complex data and generate actionable decision support should be increasingly applied to developing N-of-1 medicine--personalized medicine, where only a patient's own data is used as the basis for clinical decision support, for example, in the optimization of treatment regimens. [40,41] This approach could be advantageous in developing patient-specific therapies, where the drug and the doses are selected specifically for the given patient.…”

Section: Proposed Approachesmentioning

confidence: 99%

Overcoming Pilotitis in Digital Medicine at the Intersection of Data, Clinical Evidence, and Adoption

Egermark

Blasiak

Remus

et al. 2022

Advanced Intelligent Systems

View full text Add to dashboard Cite

The number of digital medicine technologies, a category of tools that use software and algorithmically driven products to measure or intervene to improve human health, [1] has expanded enormously in recent years, further accelerated by the need for remote care delivery during the COVID-19 pandemic, and now ranges from telemedicine to serious gaming, wearable devices, AI-supported diagnostics, and clinical decision support systems (CDSSs), among others. Building on the experiences during the pandemic, there is a real potential for the digital transformation of health care to the benefit of patients worldwide. However, the promising developments to date are at risk of stalling due to inherent limitations in the way digital medicine and especially medical artificial intelligence (AI) technologies are being developed and implemented today. [2] Already in 2017, the number of digital medicine technologies on the worldwide market exceeded 318 000, with the mobile application segment growing by over 200 daily. [3] There are widespread expectations that the surge will lead to improvement of health care globally, [4] not least evidenced by the WHO declaration on digital health, [5] and by the rapid growth of private and public investment in this field. [6,7] The advent of digital medicine bears special promise for low-and middleincome countries (LMIC), where applications such as telemedicine and CDSS may enable these countries to "leapfrog" from today's often insufficient healthcare systems into more efficient, accessible, and patient-centric models, [8] similar to what has been seen in Africa's finance sector. [9] The level of improvement possible would correspond to the digital transformation seen in other sectors of society, such as media and retail industries, with the digitalization of processes and service models, beyond the mere digitization of traditional healthcare practices seen so

show abstract

“…Thon et al [39] presented a comprehensive review of AI in process industry and provided a perspective that 'AI will take an integral role in most-if not all-fields including process and chemical engineering'. Pan et al [40] advocated for integrating simulation, ML, and statistics in data-centric engineering to create digital twins of chemical engineering systems. Further, they identified barriers in democratizing software solutions and upskilling industry practitioners as challenges for ML adoption by chemical industries.…”

Section: Fieldscalementioning

confidence: 99%

Chemical engineering at crossroads

Nandakumar

Tyagi

et al. 2022

Can J Chem Eng

View full text Add to dashboard Cite

Through periodic introspection and assessment, the chemical engineering field has developed a mature undergraduate curriculum built on a strong science background in mathematics, physics, and chemistry. This brings a unique set of skills in transport, reaction engineering, and thermodynamics, coupled with suitable process systems engineering and process design courses, to supply well-trained engineers to a vast array of process manufacturing facilities. These facilities produce basic chemicals, pharmaceuticals, oil and gas, petrochemicals, food and agricultural products, minerals, and materials. While this maturity has served existing industries well, we argue that the chemical engineering field is at crossroads between managing the curriculum of undergraduate and graduate education to supply the needs of established industries while creating innovators for emerging industries. While this is a great opportunity for yet another introspection, we caution that the inadvertent cannibalization of the field must be avoided. We do argue in favour of adding a biology sequence and a computational science sequence to the core at the undergraduate level in a related perspective article.

show abstract

Data-centric Engineering: integrating simulation, machine learning and statistics. Challenges and opportunities

Cited by 32 publications

References 54 publications

Data-centric machine learning in quantum information science

Data-centric machine learning in quantum information science

Overcoming Pilotitis in Digital Medicine at the Intersection of Data, Clinical Evidence, and Adoption

Chemical engineering at crossroads

Contact Info

Product

Resources

About