Search

Your search keyword '"Zhai, Zhaoyu"' showing total 41 results
Start Over Author "Zhai, Zhaoyu" Search Limiters Available in Library Collection
41 results on '"Zhai, Zhaoyu"'

6. Identification of Fish Hunger Degree with Deformable Attention Transformer.

Author

Wu, Yuqiang, Xu, Huanliang, Wu, Xuehui, Wang, Haiqing, and Zhai, Zhaoyu

Subjects
CONVOLUTIONAL neural networks, IDENTIFICATION of fishes, TRANSFORMER models, HUNGER, LARGEMOUTH bass, COMPUTER vision, FEATURE extraction
Abstract

Feeding is a critical process in aquaculture, as it has a direct impact on the quantity and quality of fish. With advances in convolutional neural network (CNN) and vision transformer (ViT), intelligent feeding has been widely adopted in aquaculture, as the real-time monitoring of fish behavior can lead to better feeding decisions. However, existing models still have the problem of insufficient accuracy in the fish behavior-recognition task. In this study, the largemouth bass (Micropterus salmoides) was selected as the research subject, and three categories (weakly, moderately, and strongly hungry) were defined. We applied the deformable attention to the vision transformer (DeformAtt-ViT) to identify the fish hunger degree. The deformable attention module was extremely powerful in feature extraction because it improved the fixed geometric structure of the receptive fields with data-dependent sparse attention, thereby guiding the model to focus on more important regions. In the experiment, the proposed DeformAtt-ViT was compared with the state-of-the-art transformers. Among them, DeformAtt-ViT achieved optimal performance in terms of accuracy, F1-score, recall, and precision at 95.50%, 94.13%, 95.87%, and 92.45%, respectively. Moreover, a comparative evaluation between DeformAtt-ViT and CNNs was conducted, and DeformAtt-ViT still dominated the others. We further visualized the important pixels that contributed the most to the classification result, enabling the interpretability of the model. As a prerequisite for determining the feed time, the proposed DeformAtt-ViT could identify the aggregation level of the fish and then trigger the feeding machine to be turned on. Also, the feeding machine will stop working when the aggregation disappears. Conclusively, this study was of great significance, as it explored the field of intelligent feeding in aquaculture, enabling precise feeding at a proper time. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

7. DiSignAtlas: an atlas of human and mouse disease signatures based on bulk and single-cell transcriptomics

Author

Zhai, Zhaoyu, primary, Lin, Zhewei, additional, Meng, Xuehang, additional, Zheng, Xiao, additional, Du, Yujia, additional, Li, Zhi, additional, Zhang, Xuelu, additional, Liu, Chang, additional, Zhou, Lu, additional, Zhang, Xu, additional, Tian, Zhihao, additional, Ma, Qinfeng, additional, Li, Jinhao, additional, Li, Qiang, additional, and Pan, Jianbo, additional

Published
2023
Full Text
View/download PDF

9. NAGNE: Node-to-Attribute Generation Network Embedding for Heterogeneous Network.

Author

Zhang, Zheding, Xu, Huanliang, Li, Yanbin, Zhai, Zhaoyu, and Ding, Yu

Subjects
VIRTUAL networks, DATA mining, RANDOM walks, MACHINE learning, PRIOR learning
Abstract

Heterogeneous network embedding aims to project multiple types of nodes into a low-dimensional space, and has become increasingly ubiquitous. However, several challenges have not been addressed so far. First, existing heterogeneous network embedding techniques typically rely on meta-paths to deal with the complex heterogeneous network. Using these meta-paths requires prior knowledge from domain experts for optimal meta-path selection. Second, few existing models can effectively consider both heterogeneous structural information and heterogeneous node attribute information. Third, existing models preserve the structure information by considering the first- and/or the second-order proximities, which cannot capture long-range structural information. To address these limitations, we propose a novel attributed heterogeneous network embedding model referred to as Node-to-Attribute Generation Network Embedding (NAGNE). NAGNE comprises two major components, the attributed random walk which samples node sequences in attributed heterogeneous network, and the node-to-attribute generation which learns the mapping that translates each node sequence itself from the node sequence to the node attribute sequence. Extensive experiments on three heterogeneous network datasets demonstrate that NAGNE outperforms state-of-the-art baselines in various data mining tasks. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

15. ResViT-Rice: A Deep Learning Model Combining Residual Module and Transformer Encoder for Accurate Detection of Rice Diseases.

Author

Zhang, Yujia, Zhong, Luteng, Ding, Yu, Yu, Hongfeng, and Zhai, Zhaoyu

Subjects
RICE diseases & pests, DEEP learning, CONVOLUTIONAL neural networks, RICE quality, HYBRID rice
Abstract

Rice is a staple food for over half of the global population, but it faces significant yield losses: up to 52% due to leaf blast disease and brown spot diseases, respectively. This study aimed at proposing a hybrid architecture, namely ResViT-Rice, by taking advantage of both CNN and transformer for accurate detection of leaf blast and brown spot diseases. We employed ResNet as the backbone network to establish a detection model and introduced the encoder component from the transformer architecture. The convolutional block attention module was also integrated to ResViT-Rice to further enhance the feature-extraction ability. We processed 1648 training and 104 testing images for two diseases and the healthy class. To verify the effectiveness of the proposed ResViT-Rice, we conducted comparative evaluation with popular deep learning models. The experimental result suggested that ResViT-Rice achieved promising results in the rice disease-detection task, with the highest accuracy reaching 0.9904. The corresponding precision, recall, and F1-score were all over 0.96, with an AUC of up to 0.9987, and the corresponding loss rate was 0.0042. In conclusion, the proposed ResViT-Rice can better extract features of different rice diseases, thereby providing a more accurate and robust classification output. [ABSTRACT FROM AUTHOR]

Published
2023
Full Text
View/download PDF

17. Additional file 1 of Applying spectral fractal dimension index to predict the SPAD value of rice leaves under bacterial blight disease stress

Author

Cao, YiFei, Xu, Huanliang, Song, Jin, Yang, Yao, Hu, Xiaohui, Wiyao, Korohou Tchalla, and Zhai, Zhaoyu

Abstract

Additional file 1: Figure S1. a Cultivation of the pathogenic bacteria Xanthomonas oryzae pv. Oryzae. b Infection of the bacterial blight. Figure S2. a Original reflectance curve. b SG smoothed reflectance curve. Figure S3. Measurement of SPAD using the SPAD-502 meter. Table S1. Main components of the imaging system and parameter settings. Table S2. Imaging parameters used in this experiment. Table S3. Specifications of the SPAD-502 meter.

Published
2022
Full Text
View/download PDF

23. Contributions to Case-Based Reasoning Enabled Decision Support System for Smart Agriculture

Author

Zhai, Zhaoyu, Martínez Ortega, José Fernán, and Martínez Ortega, José-Fernán

Subjects
Telecomunicaciones
Abstract

Nowadays, high demands for food from the world-wide growing population are impacting the environment and putting many pressures on agricultural productivity. As a farming management concept, smart agriculture tries to integrate advanced technologies like Internet of Things, Artificial Intelligence, and Remote Sensing into current farming practices for the purpose of boosting productivity and improving the quality of agricultural products. The core of smart agriculture emphasizes on the use of information systems and communication technologies in the cyber-physical farm management cycle. However, farmers can hardly take advantage of collected information to make proper decisions because it is difficult to transfer the explosive amount of raw data from sensors, actuators, and networks into practical knowledge for managing farming operations. Therefore, delivering an agricultural decision support system to farmers to assist them in making evidence-based decisions is needed. The ultimate objective of this thesis is to design and implement a decision support system within the Aggregate Farming in the Cloud (AFarCloud) platform. Meanwhile, the proposed decision support system tries to overcome the current challenging problems in this topic. To achieve this objective, this thesis follows the below three research areas. The first area aims at providing a general solution for delivering an agricultural decision support system for the AFarCloud platform. An architectural proposal of the decision support system framework for managing farming operations is presented in this thesis. The proposed framework defines an algorithm manager and an algorithm toolbox. The former component is responsible to configure registered decision support algorithms, while the latter component is capable of selecting a certain algorithm to generate decision supports. The proposed framework demonstrates how smart agriculture can benefit from the support of a decision support system, and therefore assist farmers in making evidence-based decisions. The second area focuses on designing a case-based reasoning (CBR) algorithm to generate decision supports for farmers. This CBR algorithm is implemented within the framework proposed in the first research area, in particular, within the algorithm toolbox component. According to the nature of the CBR algorithm, it can be divided into five steps, including representation, retrieval, reuse, revision, and retention. In this thesis, an improved CBR algorithm is proposed to overcome the detected shortcomings of the current research work. Firstly, an associated case representation formalism is presented for enhancing the typical feature vector representation. The proposed representation formalism contains the similar and dissimilar associations between past cases, enabling to compare potential similar cases preferentially. Secondly, a triangular similarity measure is designed by taking advantage of cosine and Euclidean distance measures. For providing a precise measurement, the magnitude differences between two compared N-dimensional vectors are taken into consideration. Thirdly, a fast case retrieval algorithm is developed, enabling to determine a list of similar past cases by comparing a fewer number of cases. As a consequence, the retrieval efficiency is improved while the retrieval accuracy can be guaranteed as well. Fourthly, a learning-based approach for solution reuse and revision is studied. This reuse and revision approach tries to identify the difference between the problem part of compared cases, and then update the retrieved solution based on previous experiences. Lastly, an associated case retention approach is put forward. Apart from the typical addition and deletion strategies, the proposed retention approach also concerns to update the existed associations and generate new associations for the learned cases. By enhancing each step of the CBR loop, the proposed CBR algorithm is able to generate promising decision supports with great efficiency and accuracy. The third area considers a hybrid decision support mechanism for the AFarCloud platform. It is noted that though the improved CBR algorithm can generate a satisfied result for the most queries, it may be unable to generate the decision supports when the CBR algorithm fails to retrieve a list of similar past cases. Under this circumstance, the decision support system should start other registered algorithms to carry on the task. Therefore, for coordinating the interaction between various decision support algorithms, a mediator design pattern is adopted in this hybrid decision support mechanism. Owing to the design of the mediator component, different decision support algorithms do no need to interact with each other directly. Instead, the communication work between the algorithm manager and decision support algorithms is handled by this mediator component. This hybrid decision support mechanism is verified through a preliminary proof, considering the CBR algorithm and an artificial neural network algorithm. The result suggests that the hybrid decision support mechanism can enhance the robustness of the overall decision support system. Lastly, the proposed decision support system, along with the improved CBR algorithm, are all verified by simulation. The simulation results demonstrate that the proposal in this thesis is effective and achieves better performance than previous works. ----------RESUMEN---------- Hoy en día, la alta demanda de alimentos a consecuencia del crecimiento de la población mundial está afectando al medio ambiente y ejerciendo muchas presiones sobre la producción agrícola. La agricultura inteligente, entendida como un concepto de la gestión agrícola, intenta la integración de tecnologías avanzadas, tales como Internet de las cosas, la inteligencia artificial y la teledetección, en las prácticas agrícolas actuales, con el fin de aumentar la productividad y mejorar la calidad de los productos agrícolas. El núcleo de la agricultura inteligente hace hincapié en el uso de los sistemas de información y las tecnologías de comunicaciones en el ciclo de gestión de la granja ciberfísica. Sin embargo es difícil que los granjeros puedan aprovechar la información recopilada para tomar las decisiones adecuadas. La explosiva cantidad de datos procedentes de los sensores, actuadores y redes cuesta transformarla en un conocimiento práctico que resulte útil para la administración de las operaciones agrícolas. Por lo tanto, es necesario proporcionar a los agricultores un sistema de apoyo a las decisiones agrícolas que les ayude a tomar decisiones basadas en la evidencia. El objetivo final de esta tesis es diseñar e implementar un sistema de soporte a decisiones dentro de la plataforma Aggregate Farming in the Cloud (AFarCloud). En particular, el sistema de apoyo a decisiones propuesto trata de superar los desafíos actuales en este tema. Para lograr este objetivo, esta tesis sigue las siguientes tres áreas de investigación. La primera área tiene como objetivo proporcionar una solución general para entregar un sistema de soporte a decisiones agrícolas para la plataforma AFarCloud. En esta tesis se presenta la propuesta de un marco de referencia para la arquitectura de un sistema de soporte a decisiones para la administración de las operaciones agrícolas. Este marco define dos componentes: un administrador de algoritmos, y un grupo de herramientas de algoritmos. El primero es responsable de configurar los algoritmos de soporte a decisiones que estén registrados, mientras que el segundo es capaz de seleccionar un cierto algoritmo de entre los registrados para generar soportes a decisiones. El marco propuesto demuestra cómo la agricultura inteligente puede beneficiarse del apoyo de un sistema de soporte a decisiones, y por lo tanto, cómo ayuda a los agricultores a tomar decisiones basadas en la evidencia. La segunda área se centra en el diseño de un algoritmo de razonamiento basado en casos (CBR) para generar soporte a decisiones para los agricultores. Este algoritmo CBR se implementa dentro del marco propuesto en la primera área de investigación, en particular, en el grupo de herramientas de algoritmos. De acuerdo a su naturaleza el algoritmo CBR se puede dividir en cinco pasos, que incluyen: representación, recuperación, reutilización, revisión y retención. En esta tesis, se propone un algoritmo CBR mejorado para superar las deficiencias detectadas en el trabajo de investigación. En primer lugar, se presenta la representación formal de los casos asociados para mejorar la representación típica del vector de características. La representación formal propuesta contiene las asociaciones de similitud y diferencias entre casos pasados, permitiendo comparar preferentemente los casos similares posibles. En segundo lugar, se diseña una medida de similitud triangular aprovechando las medidas de distancia del coseno y euclidiana. Para proporcionar una medición precisa se tienen en cuenta las diferencias al comparar la magnitud entre dos vectores N-dimensionales. En tercer lugar, se desarrolla un algoritmo de recuperación rápida de casos, que permite obtener una lista de casos pasados similares comparando un número menor de casos. En consecuencia, la eficiencia de la recuperación mejora a la vez que se garantiza su precisión. En cuarto lugar se estudia un enfoque basado en el aprendizaje para la reutilización y revisión de soluciones. Este enfoque de reutilización y revisión intenta identificar la diferencia entre la parte del problema de los casos comparados, para posteriormente actualizar la solución recuperada en base a las experiencias previas. Por último, se presenta un enfoque de retención del caso asociado. Además de las estrategias típicas de adición y eliminación, el enfoque de retención propuesto se refiere también a la actualización de las asociaciones existentes y a la generación nuevas asociaciones para los casos aprendidos. Al mejorar cada uno de los pasos del ciclo CBR, el algoritmo CBR propuesto puede generar soportes prometedores para la toma de decisiones, y además hacerlo con gran eficiencia y precisión. La tercera área considera un mecanismo híbrido de soporte a decisiones para la plataforma AFarCloud. Se observa que, aunque el algoritmo CBR mejorado puede generar resultados satisfactorios durante la mayor parte del tiempo, es posible que no pueda generar soportes a decisión cuando el algoritmo CBR no pueda recuperar una lista de casos pasados similares. Bajo esta circunstancia, el sistema de soporte a decisiones debe iniciar otros algoritmos registrados para llevar a cabo la tarea. Por lo tanto, para coordinar la interacción entre varios algoritmos de soporte a decisiones, se adopta un patrón de diseño tipo mediador en este mecanismo híbrido de soporte a decisiones. Gracias al diseño del componente mediador, los diferentes algoritmos de soporte a decisiones no necesitan interactuar entre sí directamente. En cambio, este componente mediador maneja el trabajo de comunicación entre el administrador de algoritmos y los algoritmos de soporte a decisiones. Este mecanismo híbrido de soporte a decisión se verifica a través de una prueba preliminar que utiliza el algoritmo CBR y una red neuronal artificial. El resultado sugiere que el mecanismo híbrido de soporte a decisiones puede mejorar la solidez del sistema general. Por último, el sistema de soporte a decisiones propuesto, junto con el algoritmo CBR mejorado, se verifican por simulación. Los resultados de la simulación demuestran que la propuesta en esta tesis es efectiva y logra un mejor rendimiento que los trabajos anteriores.

Published
2020

24. Proposal of an Automated Mission Manager for Cooperative Autonomous Underwater Vehicles

Author

Lucas Martínez, Néstor, Martínez Ortega, José Fernán, Rodríguez Molina, Jesús, Zhai, Zhaoyu, Lucas Martínez, Néstor, Martínez Ortega, José Fernán, Rodríguez Molina, Jesús, and Zhai, Zhaoyu

Abstract

In recent years there has been an increasing interest in the use of autonomous underwater vehicles (AUVs) for ocean interventions. Typical operations imply the pre-loading of a pre-generated mission plan into the AUV before being launched. Once deployed, the AUV waits for a start command to begin the execution of the plan. An onboard mission manager is responsible for handling the events that may prevent the AUV from following the plan. This approach considers the management of the mission only at the vehicle level. However, the use of a mission-level manager in coordination with the onboard mission manager could improve the handling of exogenous events that cannot be handled fully at the vehicle level. Moreover, the use of vehicle virtualization by the mission-level manager can ease the use of older AUVs. In this paper, we propose a new mission-level manager to be run at a control station. The proposed mission manager, named Missions and Task Register and Reporter (MTRR), follows a decentralized hierarchical control pattern for self-adaptive systems, and provides a basic virtualization in regard to the AUV’s planning capabilities. The MTRR has been validated as part of the SWARMs European project. During the final trials we assessed its effectiveness and measured its performance. As a result, we have identified a strong correlation between the length of mission plan and the time required to start a mission ( ρs=0.79, n=45, p0.001 ). We have also identified a possible bottleneck when accessing the repositories for storing the information from the mission. Specifically, the average time for storing the received state vectors in the relational database represented only 18.50% of the average time required for doing so in the semantic repository.

Published
2020

25. Applying case-based reasoning and a learning-based adaptation strategy to irrigation scheduling in grape farming

Author

Zhai, Zhaoyu, Martínez Ortega, José Fernán, Lucas Martínez, Néstor, Hernández Díaz, Vicente, Zhai, Zhaoyu, Martínez Ortega, José Fernán, Lucas Martínez, Néstor, and Hernández Díaz, Vicente

Abstract

As a key part of vineyard management, irrigation of grapevines puts forward the need of scheduling water resources in a more precise and efficient way. Typically, irrigation plans are generated through the use of mathematical models. However, the unwelcoming fact is that such models usually require a massive quantity of monitored data that is often unavailable or incomplete in most developing countries. As a consequence, this paper takes advantage of advanced artificial intelligence techniques, in particular, the case-based reasoning approach, to estimate the reference evapotranspiration, and therefore, to calculate the amount of irrigation water in grape farming. For improving the current case-based reasoning approach, especially the solution revision part, this paper proposes a learning-based adaptation strategy by fully making use of the hidden information in the case base. Inspired by the feature vector representation, a revision task could be also considered as a situation and action pair. The situation part attempts to capture the difference between the target case and past cases, while the action pair aims at adapting the solution to reflect the detected difference by learning adaptation knowledge from past experiences. Two retrieval tasks are involved in the revision process. On the one hand, the first one tries to retrieve an adaptation case and evaluate the difference between the new case. On the other hand, the second retrieval task should identify a collection of past cases that shares the similar difference as detected before. By learning from how the solutions of retrieved past cases were updated to solve the adaptation case, the solution of the adaptation case can be revised accordingly based on the obtained adaptation knowledge, and therefore to solve the new case. The experiment focuses on verifying the effectiveness of the case-based reasoning approach in irrigation scheduling, and evaluating the accuracy of the learning-based adaptation strategy.

Published
2020

26. An Efficient Case Retrieval Algorithm for Agricultural Case-Based Reasoning Systems, with Consideration of Case Base Maintenance

Author

Zhai, Zhaoyu, Martínez Ortega, José Fernán, Lucas Martínez, Néstor, Xu, Huanliang, Zhai, Zhaoyu, Martínez Ortega, José Fernán, Lucas Martínez, Néstor, and Xu, Huanliang

Abstract

Case-based reasoning has considerable potential to model decision support systems for smart agriculture, assisting farmers in managing farming operations. However, with the explosive amount of sensing data, these systems may achieve poor performance in knowledge management like case retrieval and case base maintenance. Typical approaches of case retrieval have to traverse all past cases for matching similar ones, leading to low efficiency. Thus, a new case retrieval algorithm for agricultural case-based reasoning systems is proposed in this paper. At the initial stage, an association table is constructed, containing the relationships between all past cases. Afterwards, attributes of a new case are compared with an entry case. According to the similarity measurement, associated similar or dissimilar cases are then compared preferentially, instead of traversing the whole case base. The association of the new case is generated through case retrieval and added in the association table at the step of case retention. The association table is also updated when a closer relationship is detected. The experiment result demonstrates that our proposal enables rapid case retrieval with promising accuracy by comparing a fewer number of past cases. Thus, the retrieval efficiency of our proposal outperforms typical approaches.

Published
2020

27. Decision support systems for agriculture 4.0: Survey and challenges

Author

Zhai, Zhaoyu, Fernán Martínez, José, Beltrán Martínez, Victoria, Lucas Martínez, Néstor, Zhai, Zhaoyu, Fernán Martínez, José, Beltrán Martínez, Victoria, and Lucas Martínez, Néstor

Abstract

Undoubtedly, high demands for food from the world-wide growing population are impacting the environment and putting many pressures on agricultural productivity. Agriculture 4.0, as the fourth evolution in farming technology, puts forward four essential requirements: increasing productivity, allocating resources reasonably, adapting to climate change, and avoiding food waste. As advanced information systems and Internet technologies are adopted in Agriculture 4.0, enormous farming data, such as temperature, humidity, soil conditions, marketing demands, and land uses, can be collected, analyzed, and processed for assisting farmers in making appropriate decisions and obtaining higher profits. Therefore, agricultural decision support systems for Agriculture 4.0 has become a very attractive topic for the research community. The objective of this paper aims at improving future agricultural decision support systems to better serve Agriculture 4.0. In this paper, the systematic literature review technique is used to survey thirteen representative decision support systems, including their applications for agricultural mission planning, water resources management, climate change adaptation, and food waste control. Each decision support system is introduced and analyzed under a systematic manner. A comprehensive evaluation is conducted from the aspects of interoperability, scalability, accessibility, usability, etc. Based on the evaluation result, upcoming challenges of employing agricultural decision support systems in Agriculture 4.0 are detected and summarized, suggesting the development trends and highlighting direct insights for future research.

Published
2020

31. An Associated Representation Method for Defining Agricultural Cases in a Case-Based Reasoning System for Fast Case Retrieval

Author

Zhai, Zhaoyu, Martínez Ortega, José Fernán, María Victoria, Beltrán Martínez, Lucas Martínez, Néstor, Zhai, Zhaoyu, Martínez Ortega, José Fernán, María Victoria, Beltrán Martínez, and Lucas Martínez, Néstor

Abstract

As an artificial intelligence technique, case-based reasoning has considerable potential to build intelligent systems for smart agriculture, providing farmers with advice about farming operation management. A proper case representation method plays a crucial role in case-based reasoning systems. Some methods like textual, attribute-value pair, and ontological representations have been well explored by researchers. However, these methods may lead to inefficient case retrieval when a large volume of data is stored in the case base. Thus, an associated representation method is proposed in this paper for fast case retrieval. Each case is interconnected with several similar and dissimilar ones. Once a new case is reported, its features are compared with historical data by similarity measurements for identifying a relative similar past case. The similarity of associated cases is measured preferentially, instead of comparing all the cases in the case base. Experiments on case retrieval were performed between the associated case representation and traditional methods, following two criteria: the number of visited cases and retrieval accuracy. The result demonstrates that our proposal enables fast case retrieval with promising accuracy by visiting fewer past cases. In conclusion, the associated case representation method outperforms traditional methods in the aspect of retrieval efficiency.

Published
2019

32. A mission planning approach for precision farming system based on multi-objective optimization

Author

Zhai, Zhaoyu and Martínez Ortega, José Fernán

Subjects
Informática, Agricultura
Abstract

As the need for food is growing continuously, the intelligent agriculture replaces the traditional agriculture by deploying various agricultural machineries to perform the agricultural missions. However, the intelligent agriculture is not smart enough, so it requires massive of human interventions. These interventions are always based on the previous experience, leading to an imprecise control over the intelligent agriculture system, especially in mission planning and resource allocation. This paper proposes a precision farming system (PFS) and treats it as a multi-agent system (MAS) with an improved federal architecture. The machineries in this system are regarded as the agents with different functionalities. Regarding the mission planning in the PFS, it is actually a multi-objective optimization problem (MOP) and this paper defines the optimization objectives. An improved algorithm, GAPSO, is proposed to plan the established missions. During the mission planning process, each agent allocates the benefits through an auction. After receiving the bids from the agent candidates, the auction host runs the GAPSO to compute the Pareto optimal solution, which is the best strategy for mission planning and resource allocation. Then, the selected agents form an agent coalition according to the computed solution and execute the agricultural missions cooperatively. In order to verify the feasibility of proposed mission planning approach, a simulation is designed to execute an agricultural mission by several machineries. The simulation result illustrates that the mission is planned properly and the resource is allocated reasonably. The proper mission planning and resource allocation enables the intelligent agriculture to increase the productivity and decrease the cost. Meanwhile, the precise use of the pesticide decreases the harm to farming fields. The theoretical analysis and simulation is a good foundation for the further study. Once the mission planning approach applies to the real scenario, it is expected to bring high economic benefits.

Published
2018

34. A Rule-Based Reasoner for Underwater Robots Using OWL and SWRL

Author

Zhai, Zhaoyu, Martínez Ortega, José Fernán, Lucas Martínez, Néstor, Castillejo Parrilla, Pedro, Zhai, Zhaoyu, Martínez Ortega, José Fernán, Lucas Martínez, Néstor, and Castillejo Parrilla, Pedro

Abstract

Web Ontology Language (OWL) is designed to represent varied knowledge about things and the relationships of things. It is widely used to express complex models and address information heterogeneity of specific domains, such as underwater environments and robots. With the help of OWL, heterogeneous underwater robots are able to cooperate with each other by exchanging information with the same meaning and robot operators can organize the coordination easier. However, OWL has expressivity limitations on representing general rules, especially the statement “If … Then … Else …”. Fortunately, the Semantic Web Rule Language (SWRL) has strong rule representation capabilities. In this paper, we propose a rule-based reasoner for inferring and providing query services based on OWL and SWRL. SWRL rules are directly inserted into the ontologies by several steps of model transformations instead of using a specific editor. In the verification experiments, the SWRL rules were successfully and efficiently inserted into the OWL-based ontologies, obtaining completely correct query results. This rule-based reasoner is a promising approach to increase the inference capability of ontology-based models and it achieves significant contributions when semantic queries are done.

Published
2018

38. A Triangular Similarity Measure for Case Retrieval in CBR and Its Application to an Agricultural Decision Support System.

Author

Zhai, Zhaoyu, Ortega, José-Fernán Martínez, Castillejo, Pedro, and Beltran, Victoria

Subjects
*DECISION support systems, *CASE-based reasoning, *PEST control, *DECISION making, *RESEMBLANCE (Philosophy)
Abstract

Case-based reasoning has been a widely-used approach to assist humans in making decisions through four steps: retrieve, reuse, revise, and retain. Among these steps, case retrieval plays a significant role because the rest of processes cannot proceed without successfully identifying the most similar past case beforehand. Some popular methods such as angle-based and distance-based similarity measures have been well explored for case retrieval. However, these methods may match inaccurate cases under certain extreme circumstances. Thus, a triangular similarity measure is proposed to identify commonalities between cases, overcoming the drawbacks of angle-based and distance-based measures. For verifying the effectiveness and performance of the proposed measure, case-based reasoning was applied to an agricultural decision support system for pest management and 300 new cases were used for testing purposes. Once a new pest problem is reported, its attributes are compared with historical data by the proposed triangular similarity measure. Farmers can obtain quick decision support on managing pest problems by learning from the retrieved solution of the most similar past case. The experimental result shows that the proposed measure can retrieve the most similar case with an average accuracy of 91.99% and it outperforms the other measures in the aspects of accuracy and robustness. [ABSTRACT FROM AUTHOR]

Published
2019
Full Text
View/download PDF

39. DMRdb: a disease-centric Mendelian randomization database for systematically assessing causal relationships of diseases with genes, proteins, CpG sites, metabolites and other diseases.

Author

Zheng X, Tian Z, Che X, Zhang X, Xiang Y, Ge Z, Zhai Z, Ma Q, and Pan J

Abstract

Exploring the causal relationships of diseases with genes, proteins, CpG sites, metabolites and other diseases is fundamental to the life sciences. However, large-scale research using Mendelian randomization (MR) analysis is currently lacking. To address this, we introduce DMRdb (http://www.inbirg.com/DMRdb/), a disease-centric Mendelian randomization database, designed to systematically assess causal relationships of diseases with genes, proteins, CpG sites, metabolites and other diseases. The database consists of three main components: (i) 6640 high-quality disease genome-wide association studies (GWASs) from public sources that were subjected to rigorous quality filtering and standardization; (ii) over 497 billion results from MR analyses involving 6640 disease GWAS datasets, 16 238 expression quantitative trait loci (eQTLs) data, 2564 protein quantitative trait loci (pQTLs) data, 12 000 methylation quantitative trait locus (meQTLs) data and 825 metabolites data and (iii) over 380 000 causal relationship pairs from 1223 literature sources relevant to MR analyses. A user-friendly online database was developed to allow users to query, search, and download all the results. In summary, we anticipate that DMRdb will be a valuable resource for advancing our understanding of disease mechanisms and identifying new biomarkers and therapeutic targets., (© The Author(s) 2024. Published by Oxford University Press on behalf of Nucleic Acids Research.)

Published
2024
Full Text
View/download PDF

40. DC 2 Net: An Asian Soybean Rust Detection Model Based on Hyperspectral Imaging and Deep Learning.

Author

Feng J, Zhang S, Zhai Z, Yu H, and Xu H

Abstract

Asian soybean rust (ASR) is one of the major diseases that causes serious yield loss worldwide, even up to 80%. Early and accurate detection of ASR is critical to reduce economic losses. Hyperspectral imaging, combined with deep learning, has already been proved as a powerful tool to detect crop diseases. However, current deep learning models are limited to extract both spatial and spectral features in hyperspectral images due to the use of fixed geometric structure of the convolutional kernels, leading to the fact that the detection accuracy of current models remains further improvement. In this study, we proposed a deformable convolution and dilated convolution neural network (DC 2 Net) for the ASR detection. The deformable convolution module was used to extract the spatial features, while the dilated convolution module was applied to extract features from the spectral dimension. We also adopted the Shapley value and the channel attention methods to evaluate the importance of each wavelength during decision-making, thereby identifying the most contributing ones. The proposed DC 2 Net can realize early asymptomatic detection of ASR even when visual symptoms have not appeared. The results of the experiment showed that the detection performance of DC 2 Net dominated state-of-the-art methods, reaching an overall accuracy at 96.73%. Meanwhile, the experimental result suggested that the Shapley Additive exPlanations method was able to extract feature wavelengths correctly, thereby helping DC 2 Net achieve reasonable performance with less input data. The research result of this study could provide early warning of ASR outbreak in advance, even at the asymptomatic period., Competing Interests: Competing interests: The authors declare that there is no conflict of interest regarding the publication of this article., (Copyright © 2024 Jiarui Feng et al.)

Published
2024
Full Text
View/download PDF

41. An Associated Representation Method for Defining Agricultural Cases in a Case-Based Reasoning System for Fast Case Retrieval.

Author

Zhai Z, Martínez Ortega JF, Beltran V, and Lucas Martínez N

Abstract

As an artificial intelligence technique, case-based reasoning has considerable potential to build intelligent systems for smart agriculture, providing farmers with advice about farming operation management. A proper case representation method plays a crucial role in case-based reasoning systems. Some methods like textual, attribute-value pair, and ontological representations have been well explored by researchers. However, these methods may lead to inefficient case retrieval when a large volume of data is stored in the case base. Thus, an associated representation method is proposed in this paper for fast case retrieval. Each case is interconnected with several similar and dissimilar ones. Once a new case is reported, its features are compared with historical data by similarity measurements for identifying a relative similar past case. The similarity of associated cases is measured preferentially, instead of comparing all the cases in the case base. Experiments on case retrieval were performed between the associated case representation and traditional methods, following two criteria: the number of visited cases and retrieval accuracy. The result demonstrates that our proposal enables fast case retrieval with promising accuracy by visiting fewer past cases. In conclusion, the associated case representation method outperforms traditional methods in the aspect of retrieval efficiency., Competing Interests: The authors declare no conflict of interest.

Published
2019
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results