webinar register page

Elucidata: Implementing ML in Biomolecular Research: Limitations and Opportunities Panel Discussion
Owing to the complexity of human biology, making precise and viable predictions can be incredibly difficult This is where ML techniques, step in. Coupled with a domain-aware understanding of the underlying data, machine learning can show incredible promise for distilling robust and reliable prognostic indicators of various health conditions.

However, ML should not be thought of as a magic wand. Not paying attention to quality of input data can lead to the development of unreliable models in clinical settings. And in the realm of biomedical research, where datasets are frequently small, iterating on data quality is critical. Collecting and labelling a large amount of training data is time-consuming and expensive, and not always a practical solution.

To put short, running a successful ML project requires a rigorous assessment of data as well as methodology. In this roundtable discussion, the panelists will discuss some advantages and caveats to applying ML in biomedical research projects and why adopting a data-centric approach goes a long way to generating accurate predictions, even with small amounts of training data.


* Required information