-
Winter wheat yield prediction using convolutional neural networks from environmental and phenological data
Authors:
Amit Kumar Srivastava,
Nima Safaei,
Saeed Khaki,
Gina Lopez,
Wenzhi Zeng,
Frank Ewert,
Thomas Gaiser,
Jaber Rahimi
Abstract:
Crop yield forecasting depends on many interactive factors, including crop genotype, weather, soil, and management practices. This study analyzes the performance of machine learning and deep learning methods for winter wheat yield prediction using an extensive dataset of weather, soil, and crop phenology variables in 271 counties across Germany from 1999 to 2019. We proposed a Convolutional Neural…
▽ More
Crop yield forecasting depends on many interactive factors, including crop genotype, weather, soil, and management practices. This study analyzes the performance of machine learning and deep learning methods for winter wheat yield prediction using an extensive dataset of weather, soil, and crop phenology variables in 271 counties across Germany from 1999 to 2019. We proposed a Convolutional Neural Network (CNN) model, which uses a 1-dimensional convolution operation to capture the time dependencies of environmental variables. We used eight supervised machine learning models as baselines and evaluated their predictive performance using RMSE, MAE, and correlation coefficient metrics to benchmark the yield prediction results. Our findings suggested that nonlinear models such as the proposed CNN, Deep Neural Network (DNN), and XGBoost were more effective in understanding the relationship between the crop yield and input data compared to the linear models. Our proposed CNN model outperformed all other baseline models used for winter wheat yield prediction (7 to 14% lower RMSE, 3 to 15% lower MAE, and 4 to 50% higher correlation coefficient than the best performing baseline across test data). We aggregated soil moisture and meteorological features at the weekly resolution to address the seasonality of the data. We also moved beyond prediction and interpreted the outputs of our proposed CNN model using SHAP and force plots which provided key insights in explaining the yield prediction results (importance of variables by time). We found DUL, wind speed at week ten, and radiation amount at week seven as the most critical features in winter wheat yield prediction.
△ Less
Submitted 25 February, 2022; v1 submitted 4 May, 2021;
originally announced May 2021.
-
Teleportation-Assisted Optical CSIGN Gates
Authors:
Fabian Ewert,
Peter van Loock
Abstract:
Reliable entangling gates for qubits encoded in single-photon states represent a major challenge on the road to scalable quantum computing architectures based on linear optics. In this work, we present two approaches to develop high-fidelity, near-deterministic controlled-sign-shift gates based on the techniques of quantum gate teleportation. On the one hand, teleportation in a discrete-variable s…
▽ More
Reliable entangling gates for qubits encoded in single-photon states represent a major challenge on the road to scalable quantum computing architectures based on linear optics. In this work, we present two approaches to develop high-fidelity, near-deterministic controlled-sign-shift gates based on the techniques of quantum gate teleportation. On the one hand, teleportation in a discrete-variable setting, i.e., for qubits, offers unit-fidelity operations but suffers from low success probabilities. Here, we apply recent results on advanced linear optical Bell measurements to reach a near-deterministic regime. On the other hand, in the setting of continuous variables, associated with coherent states, squeezing, and, typically, Gaussian states, teleportation can be performed in a deterministic fashion, but the finite amount of squeezing implies an inevitable deformation of a teleported single-mode state. Using a new generalized form of the nonlinear-sign-shift gate for gate teleportation, we are able to achieve fidelities of the resulting CSIGN gate above 90%. A special focus is also put on a comparison of the two approaches, not only with respect to fidelity and success probability, but also in terms of resource consumption.
△ Less
Submitted 29 August, 2017;
originally announced August 2017.
-
Ultrafast Fault-Tolerant Long-Distance Quantum Communication with Static Linear Optics
Authors:
Fabian Ewert,
Peter van Loock
Abstract:
We present an in-depth analysis regarding the error resistance and optimization of our all-optical Bell measurement and ultrafast long-distance quantum communication scheme proposed in [arXiv:1503.06777]. In order to promote our previous proposal from loss- to fault-tolerance, we introduce a general and compact formalism that can also be applied to other related schemes (including non-all-optical…
▽ More
We present an in-depth analysis regarding the error resistance and optimization of our all-optical Bell measurement and ultrafast long-distance quantum communication scheme proposed in [arXiv:1503.06777]. In order to promote our previous proposal from loss- to fault-tolerance, we introduce a general and compact formalism that can also be applied to other related schemes (including non-all-optical ones such as [PRL 112, 250501]). With the help of this new representation we show that our communication protocol does not only counteract the inevitable photon loss during channel transmission, but is also able to resist common experimental errors such as Pauli-type errors (bit- and phase-flips) and detector inefficiencies (losses and dark counts). Furthermore, we demonstrate that on the physical level of photonic qubits the choice of the standard linear optical Bell measurement with its limited efficiency is optimal for our setting in the sense that, apart from their potential use in state preparation, more advanced Bell measurements yield only a small decrease in resource consumption. We devise two state generation schemes that provide the required ancillary encoded Bell states (quasi-)on-demand at every station. The schemes are either based on nonlinear optics or on linear optics with multiplexing and exhibit resource costs that scale linearly or less than quadratic with the number of photons per encoded qubit, respectively. Finally, we show that it is possible to operate our communication scheme with on-off detectors instead of employing photon-number-resolving detectors.
△ Less
Submitted 14 October, 2016;
originally announced October 2016.
-
Ultrafast Long-Distance Quantum Communication with Static Linear Optics
Authors:
Fabian Ewert,
Marcel Bergmann,
Peter van Loock
Abstract:
We propose a projection measurement onto encoded Bell states with a static network of linear optical elements. By increasing the size of the quantum error correction code, both Bell measurement efficiency and photon-loss tolerance can be made arbitrarily high at the same time. As a main application, we show that all-optical quantum communication over large distances with communication rates simila…
▽ More
We propose a projection measurement onto encoded Bell states with a static network of linear optical elements. By increasing the size of the quantum error correction code, both Bell measurement efficiency and photon-loss tolerance can be made arbitrarily high at the same time. As a main application, we show that all-optical quantum communication over large distances with communication rates similar to those of classical communication is possible solely based on local state teleportations using optical sources of encoded Bell states, fixed arrays of beam splitters, and photon detectors. As another application, generalizing state teleportation to gate teleportation for quantum computation, we find that in order to achieve universality the intrinsic loss tolerance must be sacrificed and a minimal amount of feedforward has to be added.
△ Less
Submitted 4 November, 2016; v1 submitted 23 March, 2015;
originally announced March 2015.
-
3/4-efficient Bell measurement with passive linear optics and unentangled ancillae
Authors:
Fabian Ewert,
Peter van Loock
Abstract:
It is well known that an unambiguous discrimination of the four optically encoded Bell states is possible with a probability of $50\%$ at best, when using static, passive linear optics and arbitrarily many vacuum mode ancillae. By adding unentangled single-photon ancillae, we are able to surpass this limit and reach a success probability of at least $75\%$. We discuss the error robustness of the p…
▽ More
It is well known that an unambiguous discrimination of the four optically encoded Bell states is possible with a probability of $50\%$ at best, when using static, passive linear optics and arbitrarily many vacuum mode ancillae. By adding unentangled single-photon ancillae, we are able to surpass this limit and reach a success probability of at least $75\%$. We discuss the error robustness of the proposed scheme and a generalization to reach a success probability arbitrarily close to $100\%$.
△ Less
Submitted 19 March, 2014;
originally announced March 2014.
