Suchergebnisse - "Learning with Noisy Labels in Machine Learning"

Efficient annotation reduction with active learning for computer vision-based Retail Product Recognition

Autoren: Griffioen, Niels Rankovic, Nevena Zamberlan, Federico et al.

Quelle: Journal of computational social science

Schlagwörter: Artificial intelligence, Active learning, 4609 Information systems, Semi-Supervised Learning, Annotation, Social Sciences, Active Learning, 02 engineering and technology, International trade, Fabric Defect Detection in Industrial Applications, Quantum mechanics, 7. Clean energy, Annotation effort, Industrial and Manufacturing Engineering, Learning with Noisy Labels in Machine Learning, 12. Responsible consumption, Engineering, Deep Learning, Sociology, Artificial Intelligence, Meta-Learning, Machine learning, 0202 electrical engineering, electronic engineering, information engineering, Entropy (arrow of time), Business, Retail image analysis, Active Learning in Machine Learning Research, 4701 Communication and media studies, Physics, 4. Education, Retraining, 4410 Sociology, Social Sciences, Mathematical Methods, Active learning (machine learning), 15. Life on land, Computer science, Algorithm, Computer Science, Physical Sciences, Instance segmentation, Mathematics, Mathematical Methods In Social Sciences, Robust Learning

Zugangs-URL: https://research.tilburguniversity.edu/en/publications/544d20d5-2b8e-4483-b9dd-b2be4d161c17
https://doi.org/10.1007/s42001-024-00266-7
https://lirias.kuleuven.be/handle/20.500.12942/744482
https://doi.org/10.1007/s42001-024-00266-7
https://repository.uantwerpen.be/docstore/d:irua:23253
https://hdl.handle.net/10067/2054360151162165141

On data selection for the energy efficiency of neural networks: Towards a new solution based on a dynamic selectivity ratio

Autoren: Anais Boumendil Walid Bechkit Karima Benatchba et al.

Weitere Verfasser: Anais Boumendil Walid Bechkit Karima Benatchba et al.

Quelle: 2023 IEEE 35th International Conference on Tools with Artificial Intelligence (ICTAI). :325-332

Schlagwörter: Backpropagation Learning, Artificial neural network, Artificial intelligence, Economics, Epistemology, [INFO] Computer Science [cs], 7. Clean energy, Learning with Noisy Labels in Machine Learning, Selection (genetic algorithm), Engineering, Artificial Intelligence, Machine learning, 11. Sustainability, FOS: Mathematics, Neural Network Architectures, Efficient energy use, energy efficiency, Economic growth, Statistics, random data selection, Instance Selection, Neural Network Fundamentals and Applications, Computer science, FOS: Philosophy, ethics and religion, Philosophy, Operating system, Schedule, dynamic selectivity ratio, Electrical engineering, Computer Science, Physical Sciences, Convergence (economics), Quality (philosophy), Deep Learning in Computer Vision and Image Recognition, Computer Vision and Pattern Recognition, Energy (signal processing), Neural networks, Mathematics, Robust Learning

Dateibeschreibung: application/pdf

Zugangs-URL: https://hal.science/hal-04282114v1/document
https://hal.science/hal-04282114v1
https://doi.org/10.1109/ictai59109.2023.00054

Optimizing neural network algorithms for submerged membrane bioreactor: A comparative study of OVAT and RSM hyperparameter optimization techniques

Autoren: Syahira Ibrahim Norhaliza Abdul Wahab

Quelle: Water Science and Technology, Vol 89, Iss 7, Pp 1701-1724 (2024)

Schlagwörter: Artificial neural network, Artificial intelligence, Hyperparameter Optimization, neural network, Bioreactor, 02 engineering and technology, doe implementation, Environmental technology. Sanitary engineering, 01 natural sciences, Learning with Noisy Labels in Machine Learning, Bioreactors, Engineering, Response surface methodology, Artificial Intelligence, Machine learning, FOS: Mathematics, Computer Simulation, Swarm Intelligence Optimization Algorithms, Biology, TD1-1066, 0105 earth and related environmental sciences, Hyperparameter, Multi-Objective Optimization, computational time, submerged membrane bioreactor data, Optimization Applications, Botany, optimization techniques, Computer science, Ant Colony Optimization, 6. Clean water, Biochemical engineering, Particle Swarm Optimization, Application of Genetic Programming in Machine Learning, Computer Science, Physical Sciences, Neural Networks, Computer, 0210 nano-technology, Algorithms, Mathematics

Zugangs-URL: https://pubmed.ncbi.nlm.nih.gov/38619898
https://doaj.org/article/c8bb7ab8ab2345c4b2debbad56d53d08

BinCOA: An Efficient Binary Crayfish Optimization Algorithm for Feature Selection

Autoren: Nabila H. Shikoun Ahmed Salem Al‐Eraqi Islam S. Fathi

Quelle: IEEE Access, Vol 12, Pp 28621-28635 (2024)

Schlagwörter: Artificial intelligence, Hyperparameter Optimization, 7. Clean energy, Learning with Noisy Labels in Machine Learning, Selection (genetic algorithm), Artificial Intelligence, the crisscross strategy, FOS: Mathematics, 14. Life underwater, Feature Selection, Biology, Arithmetic, Mathematical optimization, Neural Network Fundamentals and Applications, 15. Life on land, Crayfish, Computer science, TK1-9971, feature selection (FS), Algorithm, Optimization algorithm, classification, the refracted opposition-based learning strategy, Fishery, 13. Climate action, Crayfish optimization algorithm (COA), Computer Science, Physical Sciences, Feature selection, Electrical engineering. Electronics. Nuclear engineering, Computer Vision and Pattern Recognition, Face Recognition and Dimensionality Reduction Techniques, Binary number, Mathematics

Zugangs-URL: https://doaj.org/article/e51406d1739048009d891522cc8b8336

Performance Augmentation of Base Classifiers Using Adaptive Boosting Framework for Medical Datasets

Autoren: Durr e Nayab Rehan Ullah Khan Ali Mustafa Qamar

Quelle: Applied Computational Intelligence and Soft Computing, Vol 2023 (2023)

Schlagwörter: Artificial neural network, Artificial intelligence, Support vector machine, Handling Imbalanced Data in Classification Problems, Health Professions, 02 engineering and technology, Boosting (machine learning), Pattern recognition (psychology), Bayesian probability, Learning with Noisy Labels in Machine Learning, Boosting, Bayes' theorem, Health Information Management, Artificial Intelligence, Support Vector Machines, Meta-Learning, Health Sciences, Machine learning, Decision tree, 0202 electrical engineering, electronic engineering, information engineering, Multilayer perceptron, Perceptron, Machine Learning in Healthcare and Medicine, Naive Bayes classifier, AdaBoost, QA75.5-76.95, 15. Life on land, Computer science, Random subspace method, 3. Good health, Electronic computers. Computer science, Computer Science, Physical Sciences, Classifier (UML), Robust Learning, Random forest

Dateibeschreibung: text/xhtml

Zugangs-URL: https://doaj.org/article/6c6ac4c6cb1f463a848e65d30d56be90

Survey on Machine Learning Biases and Mitigation Techniques

Autoren: Sunzida Siddique Mohd Ariful Haque Roy George et al.

Quelle: Digital, Vol 4, Iss 1, Pp 1-68 (2023)

Schlagwörter: Artificial intelligence, bias, Explainable Artificial Intelligence, Resource (disambiguation), Variety (cybernetics), 02 engineering and technology, Adversarial Robustness in Deep Learning Models, Adversarial system, 7. Clean energy, Learning with Noisy Labels in Machine Learning, Data science, Machine Learning, Context (archaeology), Engineering, Artificial Intelligence, Meta-Learning, Machine learning, fairness constraints, 0202 electrical engineering, electronic engineering, information engineering, FOS: Mathematics, Pathology, Risk analysis (engineering), 10. No inequality, Biology, Selection bias, pre-processing, Computer network, 4. Education, Statistics, Paleontology, QA75.5-76.95, Computer science, Management science, Programming language, machine learning, Electronic computers. Computer science, Computer Science, Physical Sciences, 8. Economic growth, Data collection, mitigation techniques, Medicine, Pipeline (software), in-processing, Mathematics, Robust Learning

Zugangs-URL: https://doaj.org/article/6d141a6735f84846ad483c27550fccee

DEVELOPMENT OF AN ENHANCED C4.5 DECISION TREE ALGORITHM USING A MEMOIZED MAPREDUCE MODEL

Autoren: Florence Paul Afolayan A. Obiniyi F. Armand Donfack-Kana et al.

Quelle: FUDMA JOURNAL OF SCIENCES. 7:156-164

Schlagwörter: FOS: Computer and information sciences, Parallel computing, Hyperparameter Optimization, Volume (thermodynamics), Categorical variable, Mathematical analysis, Quantum mechanics, 7. Clean energy, Learning with Noisy Labels in Machine Learning, Artificial Intelligence, Data Mining Techniques and Applications, Machine learning, Temporal Data Mining, Decision tree, FOS: Mathematics, Data Mining, Variable (mathematics), Data mining, Physics, Decision Trees, 15. Life on land, Computer science, Algorithm, Cache, Computer Science, Physical Sciences, Computation, Tree (set theory), Mathematics, Information Systems, Robust Learning

A bio-medical snake optimizer system driven by logarithmic surviving global search for optimizing feature selection and its application for disorder recognition

Autoren: Ruba Abu Khurma Esraa Alhenawi Malik Braik et al.

Quelle: Journal of Computational Design and Engineering. 10:2361-2383

Schlagwörter: Artificial neural network, FOS: Computer and information sciences, Artificial intelligence, Hyperparameter Optimization, Computer Science - Artificial Intelligence, Feature (linguistics), Computer Vision and Pattern Recognition (cs.CV), Computer Science - Computer Vision and Pattern Recognition, Overfitting, 02 engineering and technology, Pattern recognition (psychology), Learning with Noisy Labels in Machine Learning, 03 medical and health sciences, 0302 clinical medicine, Artificial Intelligence, Biochemistry, Genetics and Molecular Biology, Meta-Learning, Support Vector Machines, Machine learning, 0202 electrical engineering, electronic engineering, information engineering, FOS: Electrical engineering, electronic engineering, information engineering, Swarm Intelligence Optimization Algorithms, Molecular Biology, Constraint Handling, Data mining, Preprocessor, Curse of dimensionality, Prediction of Protein Subcellular Localization, Optimization Applications, Image and Video Processing (eess.IV), Life Sciences, Linguistics, Electrical Engineering and Systems Science - Image and Video Processing, Computer science, FOS: Philosophy, ethics and religion, 3. Good health, Philosophy, Artificial Intelligence (cs.AI), Computer Science, Physical Sciences, Feature selection, FOS: Languages and literature, Classifier (UML)

Zugangs-URL: http://arxiv.org/abs/2404.07216

Efficient Feature Selection in High Dimensional Data Based on Enhanced Binary Chimp Optimization Algorithms and Machine Learning

Autoren: Farid Ayeche Adel Alti

Quelle: Human-Centric Intelligent Systems, Vol 3, Iss 4, Pp 558-587 (2023)

Schlagwörter: Artificial neural network, Classification accuracy, Artificial intelligence, Hyperparameter Optimization, Economics, Feature (linguistics), Information technology, 02 engineering and technology, Learning with Noisy Labels in Machine Learning, Selection (genetic algorithm), Artificial Intelligence, Sigmoid function, Computer security, Machine learning, 0202 electrical engineering, electronic engineering, information engineering, FOS: Mathematics, Swarm Intelligence Optimization Algorithms, Feature Selection, Key (lock), Economic growth, Curse of dimensionality, Global Optimization, Arithmetic, Optimization Applications, Mathematical optimization, Linguistics, QA75.5-76.95, Rate of convergence, T58.5-58.64, Dimensionality reduction, Computer science, FOS: Philosophy, ethics and religion, Algorithm, Optimization algorithm, Philosophy, Particle Swarm Optimization, Electronic computers. Computer science, BChimp, Computer Science, Physical Sciences, Convergence (economics), Feature selection, Crossover, FOS: Languages and literature, Relevancy, Computer Vision and Pattern Recognition, Face Recognition and Dimensionality Reduction Techniques, Binary number, Mathematics

Zugangs-URL: https://doaj.org/article/c1927608a7f84e6f97ca2989e2473884

A DEBIASING VARIATIONAL AUTOENCODER FOR DEFORESTATION MAPPING

Autoren: M. X. Ortega Adarme P. J. Soto Vega G. A. O. P. Costa et al.

Quelle: The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XLVIII-M-1-2023, Pp 217-223 (2023)

Schlagwörter: Dewey Decimal Classification::500 | Naturwissenschaften::550 | Geowissenschaften, Artificial neural network, Cognitive science, Technology, Artificial intelligence, Class (philosophy), 0211 other engineering and technologies, Overfitting, Ocean Engineering, 02 engineering and technology, Debiasing, 01 natural sciences, Learning with Noisy Labels in Machine Learning, Systems engineering, Anomaly Detection in High-Dimensional Data, Task (project management), Engineering, Deep Learning, Context (archaeology), Segmentation, Artificial Intelligence, Meta-Learning, Machine learning, Psychology, Applied optics. Photonics, Automatic Road Extraction from Remote Sensing Images, Data mining, Konferenzschrift, 0105 earth and related environmental sciences, Geography, Debiasing Variational Autoencoder, Deforestation Detection, Deep learning, Autoencoder, 15. Life on land, Engineering (General). Civil engineering (General), Computer science, TA1501-1820, Map Inference, Automated Machine Learning, FOS: Psychology, Archaeology, Computer Science, Physical Sciences, Semantic Segmentation, TA1-2040, Robust Learning

Dateibeschreibung: application/pdf

Zugangs-URL: https://isprs-archives.copernicus.org/articles/XLVIII-M-1-2023/217/2023/
https://doaj.org/article/c67115d9f8e04b3c8331be9d86d72ca6

Soft computing techniques for biomedical data analysis: open issues and challenges

Autoren: Essam H. Houssein Mosa E. Hosney Marwa M. Emam et al.

Quelle: Artificial Intelligence Review. 56:2599-2649

Schlagwörter: 0301 basic medicine, Artificial neural network, Artificial intelligence, Health Professions, 02 engineering and technology, Mathematical analysis, Learning with Noisy Labels in Machine Learning, Data science, 03 medical and health sciences, Cluster analysis, Health Information Management, Artificial Intelligence, Big Data Analytics, Field (mathematics), Health Sciences, Machine learning, 0202 electrical engineering, electronic engineering, information engineering, FOS: Mathematics, Pathology, Data mining, Soft computing, Machine Learning in Healthcare and Medicine, Domain (mathematical analysis), Medical diagnosis, Pure mathematics, Computer science, 3. Good health, Computational Theory and Mathematics, Computer Science, Physical Sciences, Medicine, Mathematics, Computational Methods in Drug Discovery

Fast and Informative Model Selection Using Learning Curve Cross-Validation

Autoren: Felix Mohr Jan N. van Rijn

Quelle: IEEE Transactions on Pattern Analysis and Machine Intelligence. 45:9669-9680

Schlagwörter: FOS: Computer and information sciences, Computer Science - Machine Learning, Artificial intelligence, Semi-Supervised Learning, Handling Imbalanced Data in Classification Problems, Active Learning, 02 engineering and technology, Model selection, Learning with Noisy Labels in Machine Learning, Machine Learning (cs.LG), Context (archaeology), Selection (genetic algorithm), Artificial Intelligence, Meta-Learning, Machine learning, 0202 electrical engineering, electronic engineering, information engineering, Active Learning in Machine Learning Research, Data mining, Biology, Cost-Sensitive Learning, Paleontology, Cross-validation, Computer science, Process (computing), Operating system, Computer Science, Physical Sciences, Algorithms, Learning Curve, Robust Learning, Labeled data, Training set

Zugangs-URL: https://pubmed.ncbi.nlm.nih.gov/37028368
http://arxiv.org/abs/2111.13914

The Bigger Fish: A Comparison of Meta-Learning QSAR Models on Low-Resourced Aquatic Toxicity Regression Tasks: a comparison of meta-learning QSAR models on low-resourced aquatic toxicity regression tasks

Autoren: Thalea Schlender Markus Viljanen Jan N. van Rijn et al.

Weitere Verfasser: Thalea Schlender Markus Viljanen Jan N. van Rijn et al.

Quelle: Environ Sci Technol
Environmental science & technology 57(46), 17818-17830 (2023). doi:10.1021/acs.est.3c00334

Schlagwörter: Data Analysis, Artificial intelligence, Drug Target Identification, Aquatic toxicology, Resource (disambiguation), multi-task learning, Quantitative Structure-Activity Relationship, Organic chemistry, Learning with Noisy Labels in Machine Learning, ecotoxicology, Systems engineering, Task (project management), meta-learning, Engineering, Artificial Intelligence, Biochemistry, Genetics and Molecular Biology, Meta-Learning, aquatic ecosystem, Machine learning, Applicability domain, Animals, QSAR Modeling, Molecular Biology, Computer network, Toxicity, Geography, QSAR, Fishes, 333.7, Life Sciences, Computer science, Advances in Metabolomics Research, Chemistry, Computational Theory and Mathematics, Computer Science, Physical Sciences, learning curves, Quantitative structure–activity relationship, Benchmark (surveying), Geodesy, Computational Methods in Drug Discovery, Robust Learning, Random forest

Dateibeschreibung: application/pdf

Zugangs-URL: https://pubmed.ncbi.nlm.nih.gov/37315216
https://hdl.handle.net/1887/3716971
https://publications.rwth-aachen.de/record/962496
https://hdl.handle.net/10029/626775

NLS: An accurate and yet easy-to-interpret prediction method

Autoren: Victor Coscrato Marlene Inácio Tiago Botari et al.

Quelle: Neural Networks. 162:117-130

Schlagwörter: Artificial neural network, Artificial intelligence, Explainable Artificial Intelligence, Epistemology, Leverage (statistics), Learning with Noisy Labels in Machine Learning, Machine Learning, Anomaly Detection in High-Dimensional Data, Context (archaeology), Artificial Intelligence, Black box, Machine learning, Deep neural networks, State (computer science), Humans, Model Interpretability, Biology, Paleontology, Deep learning, Computer science, Machine Learning Interpretability, FOS: Philosophy, ethics and religion, Algorithm, Philosophy, Computer Science, Physical Sciences, Neural Networks, Computer, Supervised Machine Learning, Interpretable Models, Predictive power, Algorithms, Robust Learning

Zugangs-URL: https://pubmed.ncbi.nlm.nih.gov/36905822

Offline Pre-trained Multi-agent Decision Transformer

Autoren: Linghui Meng Muning Wen Chenyang Le et al.

Quelle: Machine Intelligence Research. 20:233-248

Schlagwörter: Artificial intelligence, Marl, Reinforcement Learning Algorithms, Generalization, Offline learning, Mathematical analysis, 7. Clean energy, decision making, Learning with Noisy Labels in Machine Learning, Multi-Agent Systems, Systems engineering, Task (project management), Engineering, Artificial Intelligence, Meta-Learning, Reinforcement learning, Machine learning, FOS: Mathematics, 14. Life underwater, Biology, Transformer, 4. Education, multi-agent reinforcement learning (MARL), Paleontology, Voltage, Pre-training model, 15. Life on land, Reinforcement Learning, Computer science, Structural basin, Operating system, Multimedia, Online learning, Application of Genetic Programming in Machine Learning, Electrical engineering, Computer Science, Physical Sciences, transformer, Online and offline, offline reinforcement learning, Mathematics, Robust Learning

Zugangs-URL: https://discovery-pp.ucl.ac.uk/id/eprint/10168378/

BaCO: A Fast and Portable Bayesian Compiler Optimization Framework

Autoren: Erik Orm Hellsten Artur Souza Johannes Lenfers et al.

Quelle: Proceedings of the 28th ACM International Conference on Architectural Support for Programming Languages and Operating Systems, Volume 4. :19-42

Schlagwörter: FOS: Computer and information sciences, Parallel computing, Computer Science - Machine Learning, Artificial intelligence, Hyperparameter Optimization, Compiler, 02 engineering and technology, Bayesian probability, Learning with Noisy Labels in Machine Learning, Machine Learning (cs.LG), Artificial Intelligence, 0202 electrical engineering, electronic engineering, information engineering, Parallel Computing and Performance Optimization, GPU Computing, Computer architecture, Active Learning in Machine Learning Research, Bayesian optimization, Computer Science - Programming Languages, Computer Science - Performance, Performance Optimization, Computer science, Programming language, Performance (cs.PF), Hardware and Architecture, Optimizing compiler, Computer Science, Physical Sciences, Programming Languages (cs.PL)

Zugangs-URL: http://arxiv.org/abs/2212.11142

Student Loss: Towards the Probability Assumption in Inaccurate Supervision

Autoren: Shuo Zhang Jian-Qing Li Hamido Fujita et al.

Quelle: IEEE Transactions on Pattern Analysis and Machine Intelligence. 46:4460-4475

Schlagwörter: LB Theory and practice of education, Composite material, Artificial intelligence, Outlier Detection, Metric (unit), Robustness (evolution), Pattern recognition (psychology), Biochemistry, Gene, Learning with Noisy Labels in Machine Learning, Anomaly Detection in High-Dimensional Data, Engineering, Artificial Intelligence, Automated Analysis of Blood Cell Images, Meta-Learning, Margin (machine learning), Machine learning, Noisy Labels, Positive and Unlabeled Data, Computer science, Materials science, Chemistry, Operations management, Aggregate (composite), Computer Science, Physical Sciences, Computer Vision and Pattern Recognition, Robust Learning, Embedding

Zugangs-URL: https://pubmed.ncbi.nlm.nih.gov/38261485

When Gradient Descent Meets Derivative-Free Optimization: A Match Made in Black-Box Scenario

Autoren: Chengcheng Han Liqing Cui Renyu Zhu et al.

Quelle: Findings of the Association for Computational Linguistics: ACL 2023. :868-880

Schlagwörter: FOS: Computer and information sciences, Artificial neural network, Financial economics, Artificial intelligence, Computer Science - Artificial Intelligence, Economics, Visual Question Answering in Images and Videos, Learning with Noisy Labels in Machine Learning, Systems engineering, Task (project management), Stochastic gradient descent, Engineering, Artificial Intelligence, Gradient method, Black box, Computer Science - Computation and Language, Gradient descent, Computer science, Algorithm, Artificial Intelligence (cs.AI), Derivative (finance), Computer Science, Physical Sciences, Deep Learning in Computer Vision and Image Recognition, Computer Vision and Pattern Recognition, Computation and Language (cs.CL), Robust Learning

Zugangs-URL: http://arxiv.org/abs/2305.10013

On Enhancing Fine-Tuning for Pre-trained Language Models

Autoren: Abir Betka Zeyd Ferhat Riyadh Barka et al.

Quelle: Proceedings of ArabicNLP 2023. :405-410

Schlagwörter: Artificial intelligence, Hyperparameter Optimization, Mathematical analysis, Quantum mechanics, Learning with Noisy Labels in Machine Learning, Systems engineering, Task (project management), Engineering, Artificial Intelligence, Meta-Learning, GRASP, FOS: Mathematics, Efficient energy use, Natural Language Processing, Domain (mathematical analysis), Topic Modeling, Physics, 4. Education, Statistics, Statistical Machine Translation and Natural Language Processing, Computer science, Language model, Language Modeling, Process (computing), Programming language, Operating system, Fine-tuning, Electrical engineering, Computer Science, Physical Sciences, Energy (signal processing), Mathematics, Robust Learning

Optimizing Document Classification: Unleashing the Power of Genetic Algorithms

Autoren: Ghulam Mustafa Abid Rauf Ahmad Sami Al-Shamayleh et al.

Quelle: IEEE Access, Vol 11, Pp 83136-83149 (2023)

Schlagwörter: Artificial intelligence, Support vector machine, Feature (linguistics), Handling Imbalanced Data in Classification Problems, Vector space model, Feature vector, Mathematical analysis, Learning with Noisy Labels in Machine Learning, Word2Vector (W2V), Document classification, Selection (genetic algorithm), Artificial Intelligence, Meta-Learning, Multi-label Text Classification in Machine Learning, Machine learning, FOS: Mathematics, Information retrieval, Feature Selection, association for computing machinery (ACM), Data mining, Metadata, Document classification (DC), Domain (mathematical analysis), bag of word (BOW), machine learning (ML), term frequency (TF), Physics, Statistical classification, Linguistics, Optics, Focus (optics), Computer science, TK1-9971, FOS: Philosophy, ethics and religion, World Wide Web, Philosophy, Computer Science, Physical Sciences, Information Retrieval, Feature selection, FOS: Languages and literature, Document Categorization, Electrical engineering. Electronics. Nuclear engineering, Mathematics

Zugangs-URL: https://doaj.org/article/a0d6c925e3744845a50ab033da8e52bd