Dimensionality reduction plots are one of the valuable ‘explainable AI’ analytics tools in Alchemite™ that enable you to enrich your understanding of your data and the relationships identified by the machine learning model. An new feature enables you to dig in further.
Dimensionality reduction allows you to explore a complete dataset in a single visualisation. The plot represents a projection of the high-dimensional data (with many columns) onto an efficient two-dimensional surface, where each point represents a row in the dataset and the distance between points indicates the similarity between those samples. Clusters of points therefore show groups of samples which are similar to each other, often identifying relationships that would otherwise be missed.
A new ‘Cluster Insights’ tab now provides additional tools enabling you to identify which columns are the most similar in each cluster, deepening your insights into the data.
Get a demo
The ‘explainable AI’ analytics tools are one of many features in Alchemite™ Analytics for applying powerful machine learning capabilities to the development of chemicals, formulations, materials, and processes. If you are not currently a user, why not request a demonstration of Alchemite™ Analytics tailored to your application area?
