New software based on Artificial Intelligence helps to interpret complex data
Experimental data is often not only highly dimensional, but also noisy and full of artefacts. This makes it difficult to interpret the data. Now a team at HZB has designed software that uses self-learning neural networks to compress the data in a smart way and reconstruct a low-noise version in the next step. This enables to recognise correlations that would otherwise not be discernible. The software has now been successfully used in photon diagnostics at the FLASH free electron laser at DESY. But it is suitable for very different applications in science.
More is not always better, but sometimes a problem. With highly complex data, which have many dimensions due to their numerous parameters, correlations are often no longer recognisable. Especially since experimentally obtained data are additionally disturbed and noisy due to influences that cannot be controlled.
Helping humans to interpret the data
Now, new software based on artificial intelligence methods can help: It is a special class of neural networks (NN) that experts call "disentangled variational autoencoder network (β-VAE)". Put simply, the first NN takes care of compressing the data, while the second NN subsequently reconstructs the data. "In the process, the two NNs are trained so that the compressed form can be interpreted by humans," explains Dr Gregor Hartmann. The physicist and data scientist supervises the Joint Lab on Artificial Intelligence Methods at HZB, which is run by HZB together with the University of Kassel.
Extracting core principles without prior knowledge
Google Deepmind had already proposed to use β-VAEs in 2017. Many experts assumed that the application in the real world would be challenging, as non-linear components are difficult to disentangle. "After several years of learning how the NNs learn, it finally worked," says Hartmann. β-VAEs are able to extract the underlying core principle from data without prior knowledge.
Photon energy of FLASH determined
In the study now published, the group used the software to determine the photon energy of FLASH from single-shot photoelectron spectra. "We succeeded in extracting this information from noisy electron time-of-flight data, and much better than with conventional analysis methods," says Hartmann. Even data with detector-specific artefacts can be cleaned up this way.
A powerful tool for different problems
"The method is really good when it comes to impaired data," Hartmann emphasises. The programme is even able to reconstruct tiny signals that were not visible in the raw data. Such networks can help uncover unexpected physical effects or correlations in large experimental data sets. "AI-based intelligent data compression is a very powerful tool, not only in photon science," says Hartmann.
Now plug and play
In total, Hartmann and his team spent three years developing the software. "But now, it is more or less plug and play. We hope that soon many colleagues will come with their data and we can support them."
- Electrocatalysis with dual functionality – an overviewHybrid electrocatalysts can produce green hydrogen, for example, and valuable organic compounds simultaneously. This promises economically viable applications. However, the complex catalytic reactions involved in producing organic compounds are not yet fully understood. Modern X-ray methods at synchrotron sources such as BESSY II, enable catalyst materials and the reactions occurring on their surfaces to be analysed in real time, in situ and under real operating conditions. This provides insights that can be used for targeted optimisation. A team has now published an overview of the current state of knowledge in Nature Reviews Chemistry.
- Successful master's degree in IR thermography on solar facadesWe are delighted to congratulate our student employee Luca Raschke on successfully completing her Master's degree in Renewable Energies at the Hochschule für Technik und Wirtschaft Berlin - and with distinction!
- BESSY II: Phosphorus chains – a 1D material with 1D electronic propertiesFor the first time, a team at BESSY II has succeeded in demonstrating the one-dimensional electronic properties in phosphorus. The samples consisted of short chains of phosphorus atoms that self-organise at specific angles on a silver substrate. Through sophisticated analysis, the team was able to disentangle the contributions of these differently aligned chains. This revealed that the electronic properties of each chain are indeed one-dimensional. Calculations predict an exciting phase transition to be expected as soon as these chains are more closely packed. While material consisting of individual chains with longer distances is semiconducting, a very dense chain structure would be metallic.