(4ao) Machine Learning: Extracting Research Pathways From Nanoscale Phenomena

Every day many research papers come out to the scientific community, summarizing its conclusions as experimental data in important fields like cell biology, medicine and nanotechnology. With the amounts of information available, there is a necessity to extract significant research trends. For over 50 years, machine learning has developed algorithms in computer science and mathematics to automatically recognize complex patterns and make intelligent decisions from information. Our research philosophy looks for a synergistic effect between first principal models and machine learning tools to reveal insights in new research areas. The application of machine learning techniques provides architecture to simultaneously integrate experimental data with our traditional chemical engineering modeling. Machine learning also offers a different perspective to understand and model systems with a scalable computational platform.

Here, we present the results of this philosophy describing the growth of platinum nanoparticles under supercritical conditions using a mathematical tool called Gaussian process regression. The results describes how the understanding of the nanoscale phenomena leads to the dynamic description of the process at the manufacturing level. We conclude this presentation with future applications of machine learning algorithms for biological and medical problems.