Pub Date : 2019-07-01DOI: 10.1109/ICMLC48188.2019.8949177
Han-Yun Chen, Ching-Hung Le, Baolian Huang
Because of the rise of environmental awareness, controlling and monitoring the electricity consumption become significant. The accuracy of the prediction of electricity consumption can directly influence the efficiency of power management. If the usage status of electricity can be predicted, it will be easy to discover if there is any unusual electricity consumption. The choice of suitable models or mathematic methods will be the essential of all. Adaptive network-based fuzzy inference system combines the concept of fuzzy and neural networks. It reserves the interpretability of fuzzy inference system and the learning ability of neural networks. We applied adaptive network-based fuzzy inference system (ANFIS) with hierarchical structure on electricity consumption prediction and grey relational analysis (GRA) on the influence of each input factors. The result showed that hierarchical ANFIS did achieve the purpose we set and GRA can effectively evaluate the magnitude of relation between factors and specific output.
{"title":"Electricity Consumption Forecasting of Buildings Using Hierarchical ANFIS and GRA","authors":"Han-Yun Chen, Ching-Hung Le, Baolian Huang","doi":"10.1109/ICMLC48188.2019.8949177","DOIUrl":"https://doi.org/10.1109/ICMLC48188.2019.8949177","url":null,"abstract":"Because of the rise of environmental awareness, controlling and monitoring the electricity consumption become significant. The accuracy of the prediction of electricity consumption can directly influence the efficiency of power management. If the usage status of electricity can be predicted, it will be easy to discover if there is any unusual electricity consumption. The choice of suitable models or mathematic methods will be the essential of all. Adaptive network-based fuzzy inference system combines the concept of fuzzy and neural networks. It reserves the interpretability of fuzzy inference system and the learning ability of neural networks. We applied adaptive network-based fuzzy inference system (ANFIS) with hierarchical structure on electricity consumption prediction and grey relational analysis (GRA) on the influence of each input factors. The result showed that hierarchical ANFIS did achieve the purpose we set and GRA can effectively evaluate the magnitude of relation between factors and specific output.","PeriodicalId":221349,"journal":{"name":"2019 International Conference on Machine Learning and Cybernetics (ICMLC)","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131980691","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-07-01DOI: 10.1109/ICMLC48188.2019.8949314
An Fang
Question texts analysis is a challenging task of the fine-grained classification due to the few annotation data and unbalanced categories. The existing approaches normally assume that each word contributes the same semantic to the question text, but ignore the different meanings of the words and the dependency relations within the text. In this paper, we propose a deep neural network with multi-layer attention mechanism to capture the extended semantic features by using a dependency parsing tree, which has the capacity to spot the central components of the question. The experimental results demonstrate that our proposed model obtains substantially improvement, comparing with several competitive baselines.
{"title":"Short-Text Question Classification Based on Dependency Parsing and Attention Mechanism","authors":"An Fang","doi":"10.1109/ICMLC48188.2019.8949314","DOIUrl":"https://doi.org/10.1109/ICMLC48188.2019.8949314","url":null,"abstract":"Question texts analysis is a challenging task of the fine-grained classification due to the few annotation data and unbalanced categories. The existing approaches normally assume that each word contributes the same semantic to the question text, but ignore the different meanings of the words and the dependency relations within the text. In this paper, we propose a deep neural network with multi-layer attention mechanism to capture the extended semantic features by using a dependency parsing tree, which has the capacity to spot the central components of the question. The experimental results demonstrate that our proposed model obtains substantially improvement, comparing with several competitive baselines.","PeriodicalId":221349,"journal":{"name":"2019 International Conference on Machine Learning and Cybernetics (ICMLC)","volume":"59 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131885615","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-07-01DOI: 10.1109/ICMLC48188.2019.8949242
Kai-Wei Chen, M. Tsai
The additive manufacturing is an intelligent manufacturing technology that can quickly build a variety of complex objects with single or different functional materials. If additive manufacturing technology can be used to print mechanical structure with sensing or electronic feature, it will be able to break through the development bottleneck of a smart gripper and achieve the goal of rapid industrial development. In this study, a multi-nozzle pneumatic extrusion additive manufacturing system for printing soft and hard material structure was developed. The structure is made of a multi-material polymer which can be fabricated by using 3D printing machine. The liquid material is extruded through a tiny nozzle and then cured by a UV lighting source. The system architecture includes a CNC controller, which controls the nozzle through two stepping motors, both positive and negative pressures and curing light source are also manipulated with peripheral I/Os. A DA controller is also applied to flexibly control the air pressure for requirement of different injected flow speed. The program part is automatically executed with a numerical control software in CNC and PLC. Different pressures were set for extrusion nozzles with different materials. The G-code data was processed by Python Language and sent to the multi-nozzle pneumatic extrusion additive manufacturing system. This paper successfully printed a sandwich pad with soft and hard material structure, including double-layer material pad and three-layer material pad. A finer printing performance than a traditional FDM machine is achieved.
{"title":"Multi-Nozzle Pneumatic Extrusion Based Additive Manufacturing System for Fabricating a Sandwich Structure with Soft and Hard Material","authors":"Kai-Wei Chen, M. Tsai","doi":"10.1109/ICMLC48188.2019.8949242","DOIUrl":"https://doi.org/10.1109/ICMLC48188.2019.8949242","url":null,"abstract":"The additive manufacturing is an intelligent manufacturing technology that can quickly build a variety of complex objects with single or different functional materials. If additive manufacturing technology can be used to print mechanical structure with sensing or electronic feature, it will be able to break through the development bottleneck of a smart gripper and achieve the goal of rapid industrial development. In this study, a multi-nozzle pneumatic extrusion additive manufacturing system for printing soft and hard material structure was developed. The structure is made of a multi-material polymer which can be fabricated by using 3D printing machine. The liquid material is extruded through a tiny nozzle and then cured by a UV lighting source. The system architecture includes a CNC controller, which controls the nozzle through two stepping motors, both positive and negative pressures and curing light source are also manipulated with peripheral I/Os. A DA controller is also applied to flexibly control the air pressure for requirement of different injected flow speed. The program part is automatically executed with a numerical control software in CNC and PLC. Different pressures were set for extrusion nozzles with different materials. The G-code data was processed by Python Language and sent to the multi-nozzle pneumatic extrusion additive manufacturing system. This paper successfully printed a sandwich pad with soft and hard material structure, including double-layer material pad and three-layer material pad. A finer printing performance than a traditional FDM machine is achieved.","PeriodicalId":221349,"journal":{"name":"2019 International Conference on Machine Learning and Cybernetics (ICMLC)","volume":"33 7-8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131978695","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-07-01DOI: 10.1109/ICMLC48188.2019.8949237
Shuai Shao, Jinseok Woo, Kouhei Yamamoto, N. Kubota
In recent years, the aging population has become a major social problem. We hope to achieve health-care system for older persons through technical means. In this study, we developed an elderly health care system based on vibration sensors. By analyzing the vibrations of behavior such as walking and falling, the system can determine the current state of the elderly and send it to the robot. Experiments show that our system can estimate the behavior of the elderly with an accuracy of 89%, in which the accuracy of fall detection is 96%.
{"title":"Elderly Health Care System Based on High Precision Vibration Sensor","authors":"Shuai Shao, Jinseok Woo, Kouhei Yamamoto, N. Kubota","doi":"10.1109/ICMLC48188.2019.8949237","DOIUrl":"https://doi.org/10.1109/ICMLC48188.2019.8949237","url":null,"abstract":"In recent years, the aging population has become a major social problem. We hope to achieve health-care system for older persons through technical means. In this study, we developed an elderly health care system based on vibration sensors. By analyzing the vibrations of behavior such as walking and falling, the system can determine the current state of the elderly and send it to the robot. Experiments show that our system can estimate the behavior of the elderly with an accuracy of 89%, in which the accuracy of fall detection is 96%.","PeriodicalId":221349,"journal":{"name":"2019 International Conference on Machine Learning and Cybernetics (ICMLC)","volume":"65 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133879369","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-07-01DOI: 10.1109/ICMLC48188.2019.8949223
Mai Kondo, T. Fujita, K. Kanda, K. Maenaka
In this study, we aimed to develop a smart-cloth for workout. A flexible elastic electrode by using silicone rubber with special carbon black material, KETJENBLACK, was fabricated and tested. The flexible conductive electrode having an expansion ability of 100% or more was successfully fabricated. The compound of special carbon black, KETJENBLACK, can offer flexible electrodes and be suitable for wearable devices.
{"title":"Nanocarbon Electrode for Wearable Device With Flexible Material","authors":"Mai Kondo, T. Fujita, K. Kanda, K. Maenaka","doi":"10.1109/ICMLC48188.2019.8949223","DOIUrl":"https://doi.org/10.1109/ICMLC48188.2019.8949223","url":null,"abstract":"In this study, we aimed to develop a smart-cloth for workout. A flexible elastic electrode by using silicone rubber with special carbon black material, KETJENBLACK, was fabricated and tested. The flexible conductive electrode having an expansion ability of 100% or more was successfully fabricated. The compound of special carbon black, KETJENBLACK, can offer flexible electrodes and be suitable for wearable devices.","PeriodicalId":221349,"journal":{"name":"2019 International Conference on Machine Learning and Cybernetics (ICMLC)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126974840","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-07-01DOI: 10.1109/ICMLC48188.2019.8949228
Sufang Zhang, Jun-Hai Zhai, Bo-Jun Xie, Yan Zhan, Xin Wang
Representation learning is the base and crucial for consequential tasks, such as classification, regression, and recognition. The goal of representation learning is to automatically learning good features with deep models. Multimodal representation learning is a special representation learning, which automatically learns good features from multiple modalities, and these modalities are not independent, there are correlations and associations among modalities. Furthermore, multimodal data are usually heterogeneous. Due to the characteristics, multimodal representation learning poses many difficulties: how to combine multimodal data from heterogeneous sources; how to jointly learning features from multimodal data; how to effectively describe the correlations and associations, etc. These difficulties triggered great interest of researchers along with the upsurge of deep learning, many deep multimodal learning methods have been proposed by different researchers. In this paper, we present an overview of deep multimodal learning, especially the approaches proposed within the last decades. We provide potential readers with advances, trends and challenges, which can be very helpful to researchers in the field of machine, especially for the ones engaging in the study of multimodal deep machine learning.
{"title":"Multimodal Representation Learning: Advances, Trends and Challenges","authors":"Sufang Zhang, Jun-Hai Zhai, Bo-Jun Xie, Yan Zhan, Xin Wang","doi":"10.1109/ICMLC48188.2019.8949228","DOIUrl":"https://doi.org/10.1109/ICMLC48188.2019.8949228","url":null,"abstract":"Representation learning is the base and crucial for consequential tasks, such as classification, regression, and recognition. The goal of representation learning is to automatically learning good features with deep models. Multimodal representation learning is a special representation learning, which automatically learns good features from multiple modalities, and these modalities are not independent, there are correlations and associations among modalities. Furthermore, multimodal data are usually heterogeneous. Due to the characteristics, multimodal representation learning poses many difficulties: how to combine multimodal data from heterogeneous sources; how to jointly learning features from multimodal data; how to effectively describe the correlations and associations, etc. These difficulties triggered great interest of researchers along with the upsurge of deep learning, many deep multimodal learning methods have been proposed by different researchers. In this paper, we present an overview of deep multimodal learning, especially the approaches proposed within the last decades. We provide potential readers with advances, trends and challenges, which can be very helpful to researchers in the field of machine, especially for the ones engaging in the study of multimodal deep machine learning.","PeriodicalId":221349,"journal":{"name":"2019 International Conference on Machine Learning and Cybernetics (ICMLC)","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134044045","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-07-01DOI: 10.1109/ICMLC48188.2019.8949193
Ke-Shiuan Lynn, Chun-Ju Chen, C. Tseng, M. Cheng, Wen-Harn Pan
Liquid chromatography/mass spectrometer (LC/MS) has become one of the most popular analytical platform for metabolomics studies owing to its wide range of detectable polarity and molecular mass. However, metabolite identification remains quite costly and time-consuming in LC/MS-based metabolomics, mostly due to lower database integrity and a separated MS/MS spectra generation process. In this work, we constructed an automated, user-friendly, and freely available tool. From a peak list, the tool first groups peaks, which are usually associated with a metabolite, based on their retention time and abundance correlation across samples. In each group, different ions are annotated and the mass of the underlying metabolite is derived. Finally, the fragments are used to match with low-energy MS/MS spectra in public databases for metabolite identification. To identify metabolites without accessible MS/MS spectra, we have developed characteristic fragment and common substructure matches. Through the above approach, we anticipate facilitating the metabolite identification in LC-MS-based metabolomics studies.
{"title":"An Automated Identification Tool for LC-MS Based Metabolomics Studies","authors":"Ke-Shiuan Lynn, Chun-Ju Chen, C. Tseng, M. Cheng, Wen-Harn Pan","doi":"10.1109/ICMLC48188.2019.8949193","DOIUrl":"https://doi.org/10.1109/ICMLC48188.2019.8949193","url":null,"abstract":"Liquid chromatography/mass spectrometer (LC/MS) has become one of the most popular analytical platform for metabolomics studies owing to its wide range of detectable polarity and molecular mass. However, metabolite identification remains quite costly and time-consuming in LC/MS-based metabolomics, mostly due to lower database integrity and a separated MS/MS spectra generation process. In this work, we constructed an automated, user-friendly, and freely available tool. From a peak list, the tool first groups peaks, which are usually associated with a metabolite, based on their retention time and abundance correlation across samples. In each group, different ions are annotated and the mass of the underlying metabolite is derived. Finally, the fragments are used to match with low-energy MS/MS spectra in public databases for metabolite identification. To identify metabolites without accessible MS/MS spectra, we have developed characteristic fragment and common substructure matches. Through the above approach, we anticipate facilitating the metabolite identification in LC-MS-based metabolomics studies.","PeriodicalId":221349,"journal":{"name":"2019 International Conference on Machine Learning and Cybernetics (ICMLC)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126396501","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-07-01DOI: 10.1109/ICMLC48188.2019.8949190
T. Takeda
Most of our daily activities consist of standing, sitting, lying and walking. Above all, sitting behavior is said to account for more than half of the waking hours, and it can be said that it is directly connected to the quality of our lives. In this research, we propose a method to evaluate the user's posture from the pressure distribution measured by the cushion type seat pressure sensor. In the proposed method, a classifier based on fuzzy inference is created from pressure values obtained from 16 pressure sensors, and the difference in posture such as normal posture and humpback, and daily life operation such as reading and paperwork are classified. The experimental results show that identification is possible with an accuracy of about 87%.
{"title":"Posture Estimation Method Using Cushion Type Seat Pressure Sensor","authors":"T. Takeda","doi":"10.1109/ICMLC48188.2019.8949190","DOIUrl":"https://doi.org/10.1109/ICMLC48188.2019.8949190","url":null,"abstract":"Most of our daily activities consist of standing, sitting, lying and walking. Above all, sitting behavior is said to account for more than half of the waking hours, and it can be said that it is directly connected to the quality of our lives. In this research, we propose a method to evaluate the user's posture from the pressure distribution measured by the cushion type seat pressure sensor. In the proposed method, a classifier based on fuzzy inference is created from pressure values obtained from 16 pressure sensors, and the difference in posture such as normal posture and humpback, and daily life operation such as reading and paperwork are classified. The experimental results show that identification is possible with an accuracy of about 87%.","PeriodicalId":221349,"journal":{"name":"2019 International Conference on Machine Learning and Cybernetics (ICMLC)","volume":"261 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114342876","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-07-01DOI: 10.1109/ICMLC48188.2019.8949284
K. Nagamune, Akito Nakano
For sports such as baseball and tennis, there are actions to throw the ball and swing the racket. There are cases injured the wrist joint by repeating this action. One such injury to the wrist joint is triangular fibrocartilage complex (TFCC) injury. TFCC is a part that keeps stability on the ulnar side of the wrist joint scale. So, if the TFCC is injured, the distance between the ulna and the radius will widen due to the wrist rotation, when the injury is severe, pain occurs on the ulnar side of the wrist joint. In the current diagnosis, there is no diagnosis to evaluate the change in distance between the ulna and the radius in the wrist rotation. Therefore, in this study, to quantitatively evaluate the change of distance between the ulna and the radius in TFCC injury, we develop a system to measure the distance between the ulna and the radius in the wrist rotation.
{"title":"A Development of a System to Measure Radioulnar Distance in Wrist-Joint Rotation Using Three-Dimensional Electromagnetic Sensor","authors":"K. Nagamune, Akito Nakano","doi":"10.1109/ICMLC48188.2019.8949284","DOIUrl":"https://doi.org/10.1109/ICMLC48188.2019.8949284","url":null,"abstract":"For sports such as baseball and tennis, there are actions to throw the ball and swing the racket. There are cases injured the wrist joint by repeating this action. One such injury to the wrist joint is triangular fibrocartilage complex (TFCC) injury. TFCC is a part that keeps stability on the ulnar side of the wrist joint scale. So, if the TFCC is injured, the distance between the ulna and the radius will widen due to the wrist rotation, when the injury is severe, pain occurs on the ulnar side of the wrist joint. In the current diagnosis, there is no diagnosis to evaluate the change in distance between the ulna and the radius in the wrist rotation. Therefore, in this study, to quantitatively evaluate the change of distance between the ulna and the radius in TFCC injury, we develop a system to measure the distance between the ulna and the radius in the wrist rotation.","PeriodicalId":221349,"journal":{"name":"2019 International Conference on Machine Learning and Cybernetics (ICMLC)","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124966343","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Almost three-quarters of the underling information in the light wave field is embodied in the phase. However, the early optical detectors can only record the intensity or amplitude of the light wave field and cannot directly extract the phase information of the light wave field. Therefore, it is necessary to use the measured amplitude or strength to reconstruct the phase information of the object, this problem is denoted phase retrieval. Phase retrieval is a matter of cardinal significance in signal processing and machine learning. The phase retrieval by convex optimization algorithm is ideal but the computational complexity is high. In 2015, Candès proposed a very effective non-convex optimization algorithm-Wirtinger flow algorithm which used spectral initialization to get a better initial value and then gradient iteration to get a promised recovery effect. Subsequently, in line with the idea, a large number of variants are devised, such as: Wirtinger flow(WF), Truncated Wirtinger Flow (TWF), Truncated Amplitude Flow (TAF), Reshaped Wirtinger Flow (RWF), Incremental Truncated Wirtinger Flow (ITWF), Incremental Reshaped Wirtinger Flow (IRWF), Robust Wirtinger Flow (Robust-WF), Sparse Wirtinger Flow (SWF), Median-TWF, Median-RWF, Generalized Wirtinger Flow (GWF), Accelerated Wirtinger Flow (AWF), Thresholded Wirtinger Flow Revisited (THWFR), Thresholded Wirtinger Flow (THWF), Reweighted Wirtinger Flow (REWF), Wirtinger Flow Method With Optimal Stepsize (WFOS), Stochastic Truncated Wirtinger Flow Algorithm (STWF), Stochastic Truncated Amplitude Flow (STAF), Reweighted Amplitude Flow (RAF), Compressive Reweighted Amplitude Flow (CRAF), SPARse Truncated Amplitude flow (SPARTA) and Sparse Wirtinger Flow Algorithm with Optimal Stepsize (SWFOS), etc. This paper analyzes and summarizes these algorithms according to their characteristics such as: initialization method, step size, iteration times, sample complexity, computational complexity, etc., so that readers can intuitively and clearly see the characteristics of each algorithm. Finally, we provide the website of the source code of some algorithms, facilitate to access and use it for readers.
{"title":"Phase Retrieval via Wirtinger Flow Algorithm and Its Variants","authors":"Jian-wei Liu, Zhi Cao, Jing Liu, Xiong-lin Luo, Wei-min Li, Nobuyasu Ito, Longteng Guo","doi":"10.1109/ICMLC48188.2019.8949170","DOIUrl":"https://doi.org/10.1109/ICMLC48188.2019.8949170","url":null,"abstract":"Almost three-quarters of the underling information in the light wave field is embodied in the phase. However, the early optical detectors can only record the intensity or amplitude of the light wave field and cannot directly extract the phase information of the light wave field. Therefore, it is necessary to use the measured amplitude or strength to reconstruct the phase information of the object, this problem is denoted phase retrieval. Phase retrieval is a matter of cardinal significance in signal processing and machine learning. The phase retrieval by convex optimization algorithm is ideal but the computational complexity is high. In 2015, Candès proposed a very effective non-convex optimization algorithm-Wirtinger flow algorithm which used spectral initialization to get a better initial value and then gradient iteration to get a promised recovery effect. Subsequently, in line with the idea, a large number of variants are devised, such as: Wirtinger flow(WF), Truncated Wirtinger Flow (TWF), Truncated Amplitude Flow (TAF), Reshaped Wirtinger Flow (RWF), Incremental Truncated Wirtinger Flow (ITWF), Incremental Reshaped Wirtinger Flow (IRWF), Robust Wirtinger Flow (Robust-WF), Sparse Wirtinger Flow (SWF), Median-TWF, Median-RWF, Generalized Wirtinger Flow (GWF), Accelerated Wirtinger Flow (AWF), Thresholded Wirtinger Flow Revisited (THWFR), Thresholded Wirtinger Flow (THWF), Reweighted Wirtinger Flow (REWF), Wirtinger Flow Method With Optimal Stepsize (WFOS), Stochastic Truncated Wirtinger Flow Algorithm (STWF), Stochastic Truncated Amplitude Flow (STAF), Reweighted Amplitude Flow (RAF), Compressive Reweighted Amplitude Flow (CRAF), SPARse Truncated Amplitude flow (SPARTA) and Sparse Wirtinger Flow Algorithm with Optimal Stepsize (SWFOS), etc. This paper analyzes and summarizes these algorithms according to their characteristics such as: initialization method, step size, iteration times, sample complexity, computational complexity, etc., so that readers can intuitively and clearly see the characteristics of each algorithm. Finally, we provide the website of the source code of some algorithms, facilitate to access and use it for readers.","PeriodicalId":221349,"journal":{"name":"2019 International Conference on Machine Learning and Cybernetics (ICMLC)","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125432289","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：