-
On the Importance of Feature Representation for Flood Map** using Classical Machine Learning Approaches
Authors:
Kevin Iselborn,
Marco Stricker,
Takashi Miyamoto,
Marlon Nuske,
Andreas Dengel
Abstract:
Climate change has increased the severity and frequency of weather disasters all around the world. Flood inundation map** based on earth observation data can help in this context, by providing cheap and accurate maps depicting the area affected by a flood event to emergency-relief units in near-real-time. Building upon the recent development of the Sen1Floods11 dataset, which provides a limited…
▽ More
Climate change has increased the severity and frequency of weather disasters all around the world. Flood inundation map** based on earth observation data can help in this context, by providing cheap and accurate maps depicting the area affected by a flood event to emergency-relief units in near-real-time. Building upon the recent development of the Sen1Floods11 dataset, which provides a limited amount of hand-labeled high-quality training data, this paper evaluates the potential of five traditional machine learning approaches such as gradient boosted decision trees, support vector machines or quadratic discriminant analysis. By performing a grid-search-based hyperparameter optimization on 23 feature spaces we can show that all considered classifiers are capable of outperforming the current state-of-the-art neural network-based approaches in terms of total IoU on their best-performing feature spaces. With total and mean IoU values of 0.8751 and 0.7031 compared to 0.70 and 0.5873 as the previous best-reported results, we show that a simple gradient boosting classifier can significantly improve over deep neural network based approaches, despite using less training data. Furthermore, an analysis of the regional distribution of the Sen1Floods11 dataset reveals a problem of spatial imbalance. We show that traditional machine learning models can learn this bias and argue that modified metric evaluations are required to counter artifacts due to spatial imbalance. Lastly, a qualitative analysis shows that this pixel-wise classifier provides highly-precise surface water classifications indicating that a good choice of a feature space and pixel-wise classification can generate high-quality flood maps using optical and SAR data. We make our code publicly available at: https://github.com/DFKI-Earth-And-Space-Applications/Flood_Map**_Feature_Space_Importance
△ Less
Submitted 1 March, 2023;
originally announced March 2023.
-
Microscopic theory for the light-induced anomalous Hall effect in graphene
Authors:
S. A. Sato,
J. W. McIver,
M. Nuske,
P. Tang,
G. Jotzu,
B. Schulte,
H. Hübener,
U. De Giovannini,
L. Mathey,
M. A. Sentef,
A. Cavalleri,
A. Rubio
Abstract:
We employ a quantum Liouville equation with relaxation to model the recently observed anomalous Hall effect in graphene irradiated by an ultrafast pulse of circularly polarized light. In the weak-field regime, we demonstrate that the Hall effect originates from an asymmetric population of photocarriers in the Dirac bands. By contrast, in the strong-field regime, the system is driven into a non-equ…
▽ More
We employ a quantum Liouville equation with relaxation to model the recently observed anomalous Hall effect in graphene irradiated by an ultrafast pulse of circularly polarized light. In the weak-field regime, we demonstrate that the Hall effect originates from an asymmetric population of photocarriers in the Dirac bands. By contrast, in the strong-field regime, the system is driven into a non-equilibrium steady state that is well-described by topologically non-trivial Floquet-Bloch bands. Here, the anomalous Hall current originates from the combination of a population imbalance in these dressed bands together with a smaller anomalous velocity contribution arising from their Berry curvature. This robust and general finding enables the simulation of electrical transport from light-induced Floquet-Bloch bands in an experimentally relevant parameter regime and creates a pathway to designing ultrafast quantum devices with Floquet-engineered transport properties.
△ Less
Submitted 11 May, 2019;
originally announced May 2019.
-
Sudden-quench dynamics of Bardeen-Cooper-Schrieffer states in deep optical lattices
Authors:
Marlon Nuske,
L. Mathey,
Eite Tiesinga
Abstract:
We determine the exact dynamics of an initial Bardeen-Cooper-Schrieffer (BCS) state of ultra-cold atoms in a deep hexagonal optical lattice. The dynamical evolution is triggered by a quench of the lattice potential, such that the interaction strength $U_f$ is much larger than the hop** amplitude $J_f$. The quench initiates collective oscillations with frequency $|U_f|/(2π)$ in the momentum occup…
▽ More
We determine the exact dynamics of an initial Bardeen-Cooper-Schrieffer (BCS) state of ultra-cold atoms in a deep hexagonal optical lattice. The dynamical evolution is triggered by a quench of the lattice potential, such that the interaction strength $U_f$ is much larger than the hop** amplitude $J_f$. The quench initiates collective oscillations with frequency $|U_f|/(2π)$ in the momentum occupation numbers and imprints an oscillating phase with the same frequency on the BCS order parameter $Δ$. The oscillation frequency of $Δ$ is not reproduced by treating the time evolution in mean-field theory. In our theory, the momentum noise (i.e. density-density) correlation functions oscillate at frequency $|U_f|/2π$ as well as at its second harmonic. For a very deep lattice, with zero tunneling energy, the oscillations of momentum occupation numbers are undamped. Non-zero tunneling after the quench leads to dephasing of the different momentum modes and a subsequent dam** of the oscillations. The dam** occurs even for a finite-temperature initial BCS state, but not for a non-interacting Fermi gas. Furthermore, dam** is stronger for larger order parameter and may therefore be used as a signature of the BCS state. Finally, our theory shows that the noise correlation functions in a honeycomb lattice will develop strong anti-correlations near the Dirac point.
△ Less
Submitted 8 August, 2016; v1 submitted 2 February, 2016;
originally announced February 2016.
-
Optimization of collisional Feshbach cooling of an ultracold nondegenerate gas
Authors:
Marlon Nuske,
Eite Tiesinga,
L. Mathey
Abstract:
We optimize a collision-induced cooling process for ultracold atoms in the nondegenerate regime. It makes use of a Feshbach resonance, instead of rf radiation in evaporative cooling, to selectively expel hot atoms from a trap. Using functional minimization we analytically show that for the optimal cooling process the resonance energy must be tuned such that it linearly follows the temperature. Her…
▽ More
We optimize a collision-induced cooling process for ultracold atoms in the nondegenerate regime. It makes use of a Feshbach resonance, instead of rf radiation in evaporative cooling, to selectively expel hot atoms from a trap. Using functional minimization we analytically show that for the optimal cooling process the resonance energy must be tuned such that it linearly follows the temperature. Here, optimal cooling is defined as maximizing the phase-space density after a fixed cooling duration. The analytical results are confirmed by numerical Monte-Carlo simulations. In order to simulate more realistic experimental conditions, we show that background losses do not change our conclusions, while additional non-resonant two-body losses make a lower initial resonance energy with non-linear dependence on temperature preferable.
△ Less
Submitted 29 December, 2014;
originally announced December 2014.
