MSc & PhD Positions - Deep Learning and Infrared Imaging of Polymers
Date and Time
The Dutcher Lab at the University of Guelph is seeking qualified MSc and PhD candidates to work on the application of machine learning (ML) and artificial intelligence (AI) to the analysis of large databases of infrared (IR) spectra collected in IR microscopy images of polymers. The goal in this work is to ultimately understand the degradation and failure mechanisms of polymers used in water transport applications, in collaboration with our industrial partner HeatLink.
We are looking for applicants who are excited to contribute to the forefront of the application of ML and AI strategies to the analysis of large databases, an emerging area at the intersection of physical and data science. Our recent use of a β-variational autoencoder (β-VAE) approach is particularly promising [1-3]. In this neural network-based approach, a very large number of IR spectra are used to train an encoder that forces the input spectra through an information bottleneck. By doing this, we can identify a small number of important generative factors called latent dimensions that are responsible for most of the measured variance in the dataset. New spectra from high resolution IR images collected on our in-house, state-of-the-art Bruker LUMOS II infrared microscope can then be analyzed using the β-VAE model to classify and track the spatial distribution of different modes of degradation in the polymers and identify new features in the data. Further insights can be achieved by using dimensionally reduced features, learned by β-VAE and other approaches, as inputs into clustering (k-means, hierarchical, and density-based) and classification (support vector machines, k-nearest neighbours, and logistic regression) models.
 M. Grossutti, J. D’Amico, J. Quintal, H. MacFarlane, A. Quirk and J.R. Dutcher. Deep Learning and Infrared Spectroscopy: Representation Learning with a β-Variational Autoencoder. J. Phys. Chem. Lett. 13, 5787 (2022).
 M. Grossutti, J. D’Amico, J. Quintal, H. MacFarlane, W.C. Wareham, A. Quirk and J.R. Dutcher. Deep Generative Modeling of Infrared Images Provides Signature of Cracking in Cross-Linked Polyethylene Pipe. ACS Appl. Mater. Interfaces 15, 22532 (2023).
 J. D’Amico, M. Grossutti and J.R. Dutcher, Deep Learning Analysis of the Propagation of Stabilizing Additive Hydrolysis in a Cross-Linked Polyethylene Pipe. ACS Appl. Polym. Mater. (2023) https://doi.org/10.1021/acsapm.3c02209
Position Requirements and Expectations
- Completed or close to completing a Bachelors or Masters degree in physics, physical chemistry or a related field of physical science
- Interest and strong motivation to work at the forefront of the application of machine learning techniques to physical science data
- Strong analytical skills and the ability to think critically and creatively
- Strong problem-solving skills and work ethic
- Excellent hands-on laboratory skills including the use of advanced instrumentation
- Ability to work safely and responsibly in a laboratory
- Ability to apply sophisticated data analysis techniques to experimental data
- Ability to program in Python and work with large databases
- Ability to work effectively in a team environment
- Strong oral and written communication skills
The anticipated start date is in Fall 2024.
Interested applicants should send a cover letter, CV and the names of up to three referees to (email@example.com). In your cover letter, you should highlight your relevant previous experience and training. Review of applications will begin immediately and continue until all positions are filled. Only applicants selected for an interview will be contacted. The Dutcher Lab and the University of Guelph are committed to building a diverse and inclusive community. All qualified applicants are invited to apply, but we particularly welcome applications from individuals that identify with groups traditionally underrepresented in the physical sciences, and we will strive to hire individuals who share our commitment to equity, diversity and inclusion.