Optimization of Weight Function for (3+1)d Phonon Propagation in Weyl Fermion Sea Expressed by Clifford Algebra Using Elman Rnn and Echo State Network

Sadataka Furui¹ and Serge Dos Santos², ¹Faculty of Science and Engineering, Teikyo University, Utsunomiya, Tochigi, Japan, ²INSA Centre Val de Loire, Universite de Tours, INSERM, Imaging Brain & Neuropsychiatry iBraiN U1253 F-411034 Blois Cedex, France

ABSTRACT

We modify the lattice simulation of (3+1)D Quantum Chromo Dynamics using fixed point actions by replacing Dirac fermions to Weyl fermions expressed by biquaternions. Paths of phonons are described by the weight function of eigenfunctions. The optimization is performed by using Elman Recurrent Neural Network and the Echo State Networks. Numerical results of the two optimizations are compared. We compare in lower dimensional systems the Time-Reversal based Nonlinear Elastic Wave Spectroscopy and the theory based on Quaternion Field Theory.

Keywords

Clifford Algebra, Quaternion Field Theory, Elman Recurrent Neural Network, Echo State Network.

"Refugee Watch" - an Extensible Framework for Forecasting Climate Measures to Assist Refugee Camps Globally and Raise Awareness

Lisa Y.W. Tang and Chung Wai, University of Nottingham, Canada

ABSTRACT

Refugees worldwide suffer enormous traumas from the loss of family members, lifetime savings, homes, and other cherished valuables. As they attempt to reclaim their “normal” daily lives in refugee camps, they are deprived of basic life essentials such as food, and safe and clean water. Nutritious goods like fresh fruits are privileges, while electricity and books are luxuries. Collectively, it takes little effort to imagine the living conditions of young children in these camps. With the frequent but unpredictable occurrences of extreme weather events, maintaining self-reliance through agricultural activities around their “temporary” homes in refugee camps is increasingly difficult. The United Nations describes drought emergency demands as unprecedented, anticipating immense challenges in maintaining agricultural farms in refugee camps. Accordingly, we developed a proof-of-concept called “Refugee Watch”, a one-stop reproducible framework that helps researchers pull data on-the-fly from data servers and try out forecasting methods with real-time series data. Implemented in Streamlit, a lightweight deployment framework for web applications, our reproducible framework is accessible at https://refugee-watch.streamlit.app/ and will be made cloneable and extensible via GitHub (shared after blind review). We hope this tool will provide an entry point for researchers and non-technical social scientists to visualize climate data, assisting policymakers in designing and implementing strategies to aid refugees living in poor conditions amid climate changes.

Keywords

Refugee, weather forecast, lightweight prototyping.