Hi, welcome to my website. I'm Paula Harder, recently I was an intern with David Rolnick at Mila, working on Physics-Constrained Climate Downscaling and a visiting researcher at the University of Oxford working on aerosol model emulation with Philip Stier and Duncan Watson-Parris. I am a Ph.D. student in Computer Science at the Fraunhofer Institute for Industrial Mathematics (ITWM) and the Scientific Computing Group (SciComp) of the University of Kaiserslautern. I am supervised by Nicolas Gauger (SciComp) and Janis Keuper (ITWM). Mey key research interest is Deep Learning applications for Climate and Weather Modeling. Besides that, I worked on robustness in Deep Learning on the topic of Adversarial Attacks and am interested in any applications of ML to climate, earth, and, space science, as well as in the broad field of AI for Social Good. I am involved with the NASA/ESA Frontier Development Lab, where I worked as an ML Scientist and returned as a team lead in 2022, looking into aerosols from wildfires.
I received a master's degree in Mathematics with a specialization in Numerical Analysis from the University of Tübingen, where I also got an education in Machine Learning. During my master's I did research projects at the German Climate Computation Center and a company for simulation of electrical networks in my spare time. After graduating I worked as a Development Engineer in the automotive industry, both on software development projects as well as data science applications. When I am not working on research projects I enjoy doing hackathons and rock climbing.
Januar 2023: Invited for a talk at the UCL Workshop on AI for Sustainability.
October 2022: Three papers I was involved in got accepted at NeurIPS workshops, one at Causal ML for Impact workshop and two at the Tackling Climate Change with ML workshop.
September 2022: Our paper on Physics-Constrained Learning for Aerosol Microphysics got accepted at Environmental Data Science.
September 2022: The work with IBM and Mila on Physics-Constrained downscaling got accepted at the AAAI Fall Symposium.
August 2022: Got interviewed on our recent work on constrained super-resolution.
June 2022: Back to FDL this year, as a team lead in the aerosol challenge.
Mai 2022: Returning to the Climate Processes group at Oxford.
May 2022: Presenting our work on Physics-Constrained Learning for Aerosol Microphysics at the Climate Informatics Conference.
March 2022: Invited talk at the ECMWF Workshop on Machine Learning.
January 2022: Starting my research internship with David Rolnick at Mila.
October 2021: Got interviewed for the 25 years anniversary of the Fraunhofer Institute ITWM.
Oktober 2021: Excited to go to Oxford as a Student visitor.
September 2021: I had a great time winning the ClimateBench challenge with my team at the NOAA AI Workshop Hackathon.
August 2021: It was great to be among the 15% to be accepted at OxML Summer school this year and a pleasure to learn more about different fields and applications in advanced Machine Learning.
June 2021: Happy to announce that our paper on Detecting AutoAttack Perturbations in the Frequency Domain was accepted at the ICML Workshop for Adversarial Machine Learning
June 2021: Our paper on emulating aerosol microphysics with machine learning has been accepted to the ICML Workshop Tackling Climate Change with AI
June 2021: Very excited to work with NASA at the FDL research sprint over this summer on how to use AI for lunar exploration
May 2021: The paper on Error estimates for the Cahn–Hilliard equation with dynamic boundary conditions got accepted at the IMA Journal of NUmerical Analysis
April 2021: Our paper on Fourier detection of Adversarial Attacks got accepted at IJCNN
January 2021: With my great teams we won prizes at two coding competitions, the Climate Crisis AI hackathon and the AI for Climate hackathon
December 2020: Presentation of the NightVision paper at the Tackling Climate Change with AI NeurIPS Workshop
September 2020:I had the pleasure to join the Climate Informatics Conference as well as participate in their hackathon
July 2020: Exited about starting my PhD at the Fraunhofer Institute for Industrial Mathematics (ITWM) and University of Kaiserslautern to work on robustness in Deep Learning and applications for climate modelling