English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Datasets in the Natural Sciences are often curated with the goal of aidingscientific understanding and hence may not always be in a form that facilitatesthe application of machine learning. In this paper, we identify three trendswithin the fields of chemical reaction prediction and synthesis design thatrequire a change in direction. First, the manner in which reaction datasets aresplit into reactants and reagents encourages testing models in anunrealistically generous manner. Second, we highlight the prevalence ofmislabelled data, and suggest that the focus should be on outlier removalrather than data fitting only. Lastly, we discuss the problem of reagentprediction, in addition to reactant prediction, in order to solve the fullsynthesis design problem, highlighting the mismatch between what machinelearning solves and what a lab chemist would need. Our critiques are alsorelevant to the burgeoning field of using machine learning to accelerateprogress in experimental Natural Sciences, where datasets are often split in abiased way, are highly noisy, and contextual variables that are not evidentfrom the data strongly influence the outcome of experiments.

Dataset Bias in the Natural Sciences: A Case Study in Chemical Reaction Prediction and Synthesis Design