Anton’s work on ‘Unsupervised classification of single-molecule data with autoencoders and transfer learning’ was accepted to Machine Learning: Science and Technology.
- Unsupervised classification of single-molecule data with autoencoders and transfer learning
Anton Vladyka and Tim Albrecht.
Machine Learning: Science and Technology, 2020
Datasets from single-molecule experiments often reflect a large variety of molecular behaviour. The exploration of such datasets can be challenging, especially if knowledge about the data is limited and a priori assumptions about expected data characteristics are to be avoided. Indeed, searching for pre-defined signal characteristics is sometimes useful, but it can also lead to information loss and the introduction of expectation bias. Here, we demonstrate how Transfer Learning-enhanced dimensionality reduction can be employed to identify and quantify hidden features in single-molecule charge transport data, in an unsupervised manner. Taking advantage of open-access neural networks trained on millions of seemingly unrelated image data, our results also show how Deep Learning methodologies can readily be employed, even if the amount of problem-specific, ‘own’ data is limited.
Nashwa’s work on ‘Surface Design: Exploiting the Instability of Small Nanoparticles on Metallic Substrates’ was published to ECS Transactions. This work was in collaboration with Dr Paramaconi Rodriguez in the School of Chemistry at the University of Birmingham.
- Surface Design: Exploiting the Instability of Small Nanoparticles on Metallic Substrates
Nashwa Awais, Paramaconi Rodriguez, and Tim Albrecht.
ECS Transactions, 97, 885–892, 2020
Thiolated Au nanoparticles have been shown to undergo fast redistribution of the capping layer and subsequently of the metal core, when in contact with bare Au substrates. This is the result of an intricate interplay of entropic and enthalpic factors, which are likely affected by the choice of core metal and capping chemistry. This raises interesting questions, for example whether such a process could be used to modify and functionalize electrode substrates in a well-defined and controlled manner. Here, we report results for Pt and Au nanoparticles on Au and Pt substrates, based on the electrochemical response of the modified electrodes towards the oxidation of glycerol in alkaline media. Our study provides evidence that Pt nanoparticles on Au substrates remain relatively stable, while Au nanoparticles on Pt substrates readily decompose. This is in accordance with initial expectations based on the energetics of the thiol/metal bond.