Searcharxiv⌕ Search

arXiv subjects

Andreas Müller

Publications and source records attributed to Andreas Müller.

At least 37 records · Page 2Linked to original sources
arXiv2024

A tour d'horizon of de Casteljau's work

Whilst Paul de Casteljau is now famous for his fundamental algorithm of curve and surface approximation, little is known about his other findings. This article offers an insight into his results in geometry, algebra and number theory. Related to geometry, his classical algorithm is reviewed as an index reduction of a polar form. This idea is used to show de Casteljau's algebraic way of smoothing, which long went unnoticed. We will also see an analytic polar form and its use in finding the intersection of two curves. The article summarises unpublished material on metric geometry. It includes theoretical advances, e.g., the 14-point strophoid or a way to link Apollonian circles with confocal conics, and also practical applications such as a recurrence for conjugate mirrors in geometric optics. A view on regular polygons leads to an approximation of their diagonals by golden matrices, a generalisation of the golden ratio. Relevant algebraic findings include matrix quaternions (and anti-quaternions) and their link with Lorentz' equations. De Casteljau generalised the Euclidean algorithm and developed an automated method for approximating the roots of a class of polynomial equations. His contributions to number theory not only include aspects on the sum of four squares as in quaternions, but also a view on a particular sum of three cubes. After a review of a complete quadrilateral in a heptagon and its angles, the paper concludes with a summary of de Casteljau's key achievements. The article contains a comprehensive bibliography of de Casteljau's works, including previously unpublished material.

math.HO↗
arXiv2024

Understanding Of Size And Scale And Order-Of-Magnitude Reasoning In Secondary Science: A Teaching Experiment With Worked Examples As Educational Scaffold

Understanding size and scale (USS) and order-of-magnitude reasoning (OMR) are critical for scientific literacy. This study examines an educational approach to enhance these skills in high school science, considering the cognitive prerequisites and challenges. It explores worked examples (WEs) as an effective method to teach USS and OMR, comparing their impact on students' knowledge and transfer abilities with conventional tasks in a quasi-experimental study. The results indicate significant improvements in procedural and conceptual knowledge, effective for diverse learners regardless of prior knowledge.

physics.ed-ph↗
arXiv2024

The Impact of Uniform Inputs on Activation Sparsity and Energy-Latency Attacks in Computer Vision

Resource efficiency plays an important role for machine learning nowadays. The energy and decision latency are two critical aspects to ensure a sustainable and practical application. Unfortunately, the energy consumption and decision latency are not robust against adversaries. Researchers have recently demonstrated that attackers can compute and submit so-called sponge examples at inference time to increase the energy consumption and decision latency of neural networks. In computer vision, the proposed strategy crafts inputs with less activation sparsity which could otherwise be used to accelerate the computation. In this paper, we analyze the mechanism how these energy-latency attacks reduce activation sparsity. In particular, we find that input uniformity is a key enabler. A uniform image, that is, an image with mostly flat, uniformly colored surfaces, triggers more activations due to a specific interplay of convolution, batch normalization, and ReLU activation. Based on these insights, we propose two new simple, yet effective strategies for crafting sponge examples: sampling images from a probability distribution and identifying dense, yet inconspicuous inputs in natural datasets. We empirically examine our findings in a comprehensive evaluation with multiple image classification models and show that our attack achieves the same sparsity effect as prior sponge-example methods, but at a fraction of computation effort. We also show that our sponge examples transfer between different neural networks. Finally, we discuss applications of our findings for the good by improving efficiency by increasing sparsity.

cs.CR↗
arXiv2023

Sea Level Rise by Climate Change. An order-of-magnitude approach to an environmental problem

A figure, taken from a science literacy test, illustrates the distribution of water across various locations on Earth, represented as though the entire volume is contained in 100 buckets. Using this figure , and other basic, readily available geographic information one can deduce the approximate value of about 70m for the rise in sea level due to the melting of all Earth's land ice, a value often discussed in media and public discourse.

physics.pop-ph↗
arXiv2023

On the Detection of Image-Scaling Attacks in Machine Learning

Image scaling is an integral part of machine learning and computer vision systems. Unfortunately, this preprocessing step is vulnerable to so-called image-scaling attacks where an attacker makes unnoticeable changes to an image so that it becomes a new image after scaling. This opens up new ways for attackers to control the prediction or to improve poisoning and backdoor attacks. While effective techniques exist to prevent scaling attacks, their detection has not been rigorously studied yet. Consequently, it is currently not possible to reliably spot these attacks in practice. This paper presents the first in-depth systematization and analysis of detection methods for image-scaling attacks. We identify two general detection paradigms and derive novel methods from them that are simple in design yet significantly outperform previous work. We demonstrate the efficacy of these methods in a comprehensive evaluation with all major learning platforms and scaling algorithms. First, we show that image-scaling attacks modifying the entire scaled image can be reliably detected even under an adaptive adversary. Second, we find that our methods provide strong detection performance even if only minor parts of the image are manipulated. As a result, we can introduce a novel protection layer against image-scaling attacks.

cs.CR↗
arXiv2023

Physics lab courses under digital transformation: A tri-national survey among university lab instructors about the role of new digital technologies and learning objectives

Physics lab courses permanently undergo transformations, in recent times especially to adapt to the emergence of new digital technologies and the Covid-19 pandemic in which digital technologies facilitated distance learning. Since these transformations often occur within individual institutions, it is useful to get an overview of these developments by capturing the status quo of digital technologies and the related acquisition of digital competencies in physics lab courses. Thus, we conducted a survey among physics lab instructors (N=79) at German, Finnish, and Croatian universities. The findings reveal that lab instructors already use a variety of digital technologies and that the pandemic particularly boosted the use of smartphones/tablets, simulations, and digital tools for communication/collaboration/organization. The participants generally showed a positive attitude toward using digital technologies in physics lab courses, especially due to their potential for experiments and students' competence acquisition, motivational effects, and contemporaneity. Acquiring digital competencies is rated as less important than established learning objectives, however, collecting and processing data with digital tools was rated as an important competency that students should acquire. The instructors perceived open forms of labwork and particular digital technologies for specific learning objectives (e.g., microcontrollers for experimental skills) as useful for reaching their learning objectives. Our survey contributes to the reflection of what impact the emergence of digital technologies in our society and the Covid-19 pandemic had on physics lab courses and reveals first indications for the future transformation of hands-on university physics education.

physics.ed-ph↗
arXiv2023

A framework for designing experimental tasks in contemporary physics lab courses

While lab courses are an integral part of studying physics aiming at a huge variety of learning objectives, research has shown that typical lab courses do not reach all the desired goals. While diverse approaches by lab instructors and researchers try to increase the effectiveness of lab courses, experimental tasks remain the core of any lab course. To keep an overview of these developments and to give instructors (and researchers) a guideline for their own professional efforts at hand, we introduce a research-informed framework for designing experimental tasks in contemporary physics lab courses. In addition, we demonstrate within the scope of the EU-co-funded DigiPhysLab-project how the framework can be used to characterize existing or develop new high-quality experimental tasks for physics lab courses.

physics.ed-ph↗
arXiv2023

Detecting Spells in Fantasy Literature with a Transformer Based Artificial Intelligence

Transformer architectures and models have made significant progress in language-based tasks. In this area, is BERT one of the most widely used and freely available transformer architecture. In our work, we use BERT for context-based phrase recognition of magic spells in the Harry Potter novel series. Spells are a common part of active magic in fantasy novels. Typically, spells are used in a specific context to achieve a supernatural effect. A series of investigations were conducted to see if a Transformer architecture could recognize such phrases based on their context in the Harry Potter saga. For our studies a pre-trained BERT model was used and fine-tuned utilising different datasets and training methods to identify the searched context. By considering different approaches for sequence classification as well as token classification, it is shown that the context of spells can be recognised. According to our investigations, the examined sequence length for fine-tuning and validation of the model plays a significant role in context recognition. Based on this, we have investigated whether spells have overarching properties that allow a transfer of the neural network models to other fantasy universes as well. The application of our model showed promising results and is worth to be deepened in subsequent studies.

cs.CL↗
arXiv2023

Smartphone experiments in Undergraduate Research

Smartphones and tablets are an integral part of our daily lives, and their capabilities extend well beyond communication and entertainment. With a broad choice of built-in sensors, using these mobile devices as experimental tools (MDETs) allows for a many different measurements, covering several fields of physics (mechanics, acoustics and waves, magnetism, optics, etc.). Building on this development, the present contribution is about exploring the potential of MDETs in physics undergraduate research. Two examples related to acoustics (bottle Helmholtz resonator, singing glasses) will be discussed in detail, and four further possibilities are referred to as perspectives. Results of these student research projects are encouraging throughout (accuracy, agreement with theory), in many cases providing a basis for further improvements and insight. Additionally, it is argued that these examples provide a "proof of concept" that the use of smartphones for experimental projects can mobilize and stimulate among students the educational potential of "higher order thinking skills" (HOTs) widely discussed in the literature on undergraduate research, such as autonomy, curiosity, creativity, and others.

physics.ed-ph↗
arXiv2023

Newspaper Story Problems and Other Tasks for Context Based Physics Education: A Research Based report on classroom practice

Background: Context Based Science Education (CBSE) has a long-standing tradition and is discussed as a highly promising approach in science education. It is supposed that CBSE can foster pupils engagement and learning. However, classroom implementations of CBSE based on solid empirical evidence are surprisingly scarce. Purpose: The present research-based report of practice seeks to bridge this theory practice gap for some specific forms of CBSE. We examine the use of science problems based on newspaper articles and the real-life contexts they provide (newspaper story problems, NSP). Design and Methods: While the research background has been reported elsewhere, the main objective of the present contribution is to provide a detailed account of the practical aspects of the approach. Two concrete, curriculum-relevant classroom teaching experiments based on newspaper story problems are reported, combined with a quasi-experimental study comparing NSPs against conventional textbook problems. The implementation of the teaching learning sequence in classroom practice is described in detail. Additionally, tasks types similar to NSPs, but using other ways of contextualisation (e.g. advertisements) will be discussed as perspective. Results: A considerable improvement in motivation was found, which proved stable at least in the medium term. Learning turned out to also be fostered to a sizeable extent, including the educationally important issue of transfer. Conclusions: The use of NSPs as a form of CBSE can have large positive, robust, and sustainable effects of both motivation and learning. Being flexible and practical to implement, they appear thus highly suited to classroom application. In perspective, a series of similar forms of tasks implementing CBSE is presented, such as by experimental of aesthetic contexts.

physics.ed-ph↗
arXiv2023

Listen! A smartphone inquiry on the domino effect

In this work, we investigate the phenomenon of a chain of falling dominoes ("domino effect") by an acoustical measurement with a smartphone. Specifically, we will present an approximate model valid for a considerable range of realistic parameters of the domino effect; a new, easy and low-cost experimental method using acoustical data captured by smartphones; and a comparison of these experimental data to the simplified treatment and to other, more comprehensive theories. The results of the experiments show good agreement with other measurements and theory, both advanced ones from the literature, and our own simplified treatment: qualitatively the approach to a constant propagation velocity of the domino front after a small number of toppled dominoes as predicted by theory; quantitatively the numerical value of the asymptotic propagation velocity. The example can be seen as a demonstration of the use of mobile devices as experimental tools for student projects about a phenomenon of "everyday physics", and in the same time of interest for current research (the domino effect as a paradigm of collective dynamics with applications e.g. in neuronal signal transduction.

physics.ed-ph↗
arXiv2023

Organelle-specific segmentation, spatial analysis, and visualization of volume electron microscopy datasets

Volume electron microscopy is the method of choice for the in-situ interrogation of cellular ultrastructure at the nanometer scale. Recent technical advances have led to a rapid increase in large raw image datasets that require computational strategies for segmentation and spatial analysis. In this protocol, we describe a practical and annotation-efficient pipeline for organelle-specific segmentation, spatial analysis, and visualization of large volume electron microscopy datasets using freely available, user-friendly software tools that can be run on a single standard workstation. We specifically target researchers in the life sciences with limited computational expertise, who face the following tasks within their volume electron microscopy projects: i) How to generate 3D segmentation labels for different types of cell organelles while minimizing manual annotation efforts, ii) how to analyze the spatial interactions between organelle instances, and iii) how to best visualize the 3D segmentation results. To meet these demands we give detailed guidelines for choosing the most efficient segmentation tools for the specific cell organelle. We furthermore provide easily executable components for spatial analysis and 3D rendering and bridge compatibility issues between freely available open-source tools, such that others can replicate our full pipeline starting from a raw dataset up to the final plots and rendered images. We believe that our detailed description can serve as a valuable reference for similar projects requiring special strategies for single- or multiple organelle analysis which can be achieved with computational resources commonly available to single-user setups.

cs.CV↗
arXiv2022

LlamaTune: Sample-Efficient DBMS Configuration Tuning

Tuning a database system to achieve optimal performance on a given workload is a long-standing problem in the database community. A number of recent works have leveraged ML-based approaches to guide the sampling of large parameter spaces (hundreds of tuning knobs) in search for high performance configurations. Looking at Microsoft production services operating millions of databases, sample efficiency emerged as a crucial requirement to use tuners on diverse workloads. This motivates our investigation in LlamaTune, a tuner design that leverages domain knowledge to improve the sample efficiency of existing optimizers. LlamaTune employs an automated dimensionality reduction technique based on randomized projections, a biased-sampling approach to handle special values for certain knobs, and knob values bucketization, to reduce the size of the search space. LlamaTune compares favorably with the state-of-the-art optimizers across a diverse set of workloads. It identifies the best performing configurations with up to $11\times$ fewer workload runs, and reaching up to $21\%$ higher throughput. We also show that benefits from LlamaTune generalize across both BO-based and RL-based optimizers, as well as different DBMS versions. While the journey to perform database tuning at cloud-scale remains long, LlamaTune goes a long way in making automatic DBMS tuning practical at scale.

cs.DB↗
arXiv2021

OpenML-Python: an extensible Python API for OpenML

OpenML is an online platform for open science collaboration in machine learning, used to share datasets and results of machine learning experiments. In this paper we introduce OpenML-Python, a client API for Python, opening up the OpenML platform for a wide range of Python-based tools. It provides easy access to all datasets, tasks and experiments on OpenML from within Python. It also provides functionality to conduct machine learning experiments, upload the results to OpenML, and reproduce results which are stored on OpenML. Furthermore, it comes with a scikit-learn plugin and a plugin mechanism to easily integrate other machine learning libraries written in Python into the OpenML ecosystem. Source code and documentation is available at https://github.com/openml/openml-python/.

cs.LG↗
arXiv2021

Learning Multiple Defaults for Machine Learning Algorithms

The performance of modern machine learning methods highly depends on their hyperparameter configurations. One simple way of selecting a configuration is to use default settings, often proposed along with the publication and implementation of a new algorithm. Those default values are usually chosen in an ad-hoc manner to work good enough on a wide variety of datasets. To address this problem, different automatic hyperparameter configuration algorithms have been proposed, which select an optimal configuration per dataset. This principled approach usually improves performance but adds additional algorithmic complexity and computational costs to the training procedure. As an alternative to this, we propose learning a set of complementary default values from a large database of prior empirical results. Selecting an appropriate configuration on a new dataset then requires only a simple, efficient and embarrassingly parallel search over this set. We demonstrate the effectiveness and efficiency of the approach we propose in comparison to random search and Bayesian Optimization.

stat.ML↗
arXiv2021

Spin wave localization and guiding by magnon band structure engineering in yttrium iron garnet

In spintronics the propagation of spin-wave excitations in magnetically ordered materials can also be used to transport and process information. One of the most popular materials in this regard is the ferrimagnetic insulator yttrium-iron-garnet due its exceptionally small spin-wave damping parameter. While the small relaxation rate allows for large propagation length of magnetic excitations, it also leads to non-locality of the magnetic properties. By imaging spin waves their band structure is mapped. In doing so wave vector selection is shown to suppress dispersion effects to a large extent allowing for local measurements of spin relaxation. Moreover we demonstrate even higher control of magnon propagation by employing the wave vector selectivity near an avoided crossing of different spin-wave modes where the group velocity approaches zero. Here local engineering of the dispersion allows constructing magnonic waveguides and at the same time reveals the local relaxation properties.

cond-mat.mes-hall↗
arXiv2020

On The Gaussian Approximation To Bayesian Posterior Distributions

The present article derives the minimal number $N$ of observations needed to consider a Bayesian posterior distribution as Gaussian. Two examples are presented. Within one of them, a chi-squared distribution, the observable $x$ as well as the parameter $ξ$ are defined all over the real axis, in the other one, the binomial distribution, the observable $x$ is an entire number while the parameter $ξ$ is defined on a finite interval of the real axis. The required minimal $N$ is high in the first case and low for the binomial model. In both cases the precise definition of the measure $μ$ on the scale of $ξ$ is crucial.

math.ST↗