Multilabel learning is an important topic in machine learning research. Evaluating models in multilabel settings requires specific cross validation methods designed for multilabel data. In this article, we show that the most widely used cross validation split quality measure does not behave adequately with multilabel data that has strong class imbalance. We present improved measures and an algorithm, optisplit, for optimizing cross validations splits. Extensive comparison of various types of cross validation methods shows that optisplit produces more even cross validation splits than the existing methods and it is among the fastest methods with good splitting performance.
{"title":"Novel split quality measures for stratified multilabel cross validation with application to large and sparse gene ontology datasets","authors":"Henri Tiittanen, L. Holm, Petri Toronen","doi":"10.3934/aci.222003","DOIUrl":"https://doi.org/10.3934/aci.222003","url":null,"abstract":"Multilabel learning is an important topic in machine learning research. Evaluating models in multilabel settings requires specific cross validation methods designed for multilabel data. In this article, we show that the most widely used cross validation split quality measure does not behave adequately with multilabel data that has strong class imbalance. We present improved measures and an algorithm, optisplit, for optimizing cross validations splits. Extensive comparison of various types of cross validation methods shows that optisplit produces more even cross validation splits than the existing methods and it is among the fastest methods with good splitting performance.","PeriodicalId":414924,"journal":{"name":"Applied Computing and Intelligence","volume":"75 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124197773","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}
CNN models already play an important role in classification of crop and weed with high accuracy, more than 95% as reported in literature. However, to manually choose and fine-tune the deep learning models becomes laborious and indispensable in most traditional practices and research. Moreover, the classic objective functions are not thoroughly compatible with agricultural farming tasks as the corresponding models suffer from misclassifying crop to weed, often more likely than in other deep learning application domains. In this paper, we applied autonomous machine learning with a new objective function for crop and weed classification, achieving higher accuracy and lower crop killing rate (rate of identifying a crop as a weed). The experimental results show that our method outperforms state-of-the-art applications, for example, ResNet and VGG19.
{"title":"Crop and weed classification based on AutoML","authors":"Xuetao Jiang, Binbin Yong, Soheila Garshasbi, Jun Shen, Meiyu Jiang, Qingguo Zhou","doi":"10.3934/aci.2021003","DOIUrl":"https://doi.org/10.3934/aci.2021003","url":null,"abstract":"\u0000 CNN models already play an important role in classification of crop and weed with high accuracy, more than 95% as reported in literature. However, to manually choose and fine-tune the deep learning models becomes laborious and indispensable in most traditional practices and research. Moreover, the classic objective functions are not thoroughly compatible with agricultural farming tasks as the corresponding models suffer from misclassifying crop to weed, often more likely than in other deep learning application domains. In this paper, we applied autonomous machine learning with a new objective function for crop and weed classification, achieving higher accuracy and lower crop killing rate (rate of identifying a crop as a weed). The experimental results show that our method outperforms state-of-the-art applications, for example, ResNet and VGG19.\u0000","PeriodicalId":414924,"journal":{"name":"Applied Computing and Intelligence","volume":"77 2-3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133088062","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}
In clustering, the cost function that is commonly used involves calculating all-pairwise squared distances. In this paper, we formulate the cost function using mean squared error and show that this leads to more balanced clustering compared to centroid-based distance functions, like the sum of squared distances in $ k $-means. The clustering method has been formulated as a cut-based approach, more intuitively called Squared cut (Scut). We introduce an algorithm for the problem which is faster than the existing one based on the Stirling approximation. Our algorithm is a sequential variant of a local search algorithm. We show by experiments that the proposed approach provides better overall optimization of both mean squared error and cluster balance compared to existing methods.
{"title":"All-pairwise squared distances lead to more balanced clustering","authors":"Mikko I. Malinen, P. Fränti","doi":"10.3934/aci.2023006","DOIUrl":"https://doi.org/10.3934/aci.2023006","url":null,"abstract":"In clustering, the cost function that is commonly used involves calculating all-pairwise squared distances. In this paper, we formulate the cost function using mean squared error and show that this leads to more balanced clustering compared to centroid-based distance functions, like the sum of squared distances in $ k $-means. The clustering method has been formulated as a cut-based approach, more intuitively called Squared cut (Scut). We introduce an algorithm for the problem which is faster than the existing one based on the Stirling approximation. Our algorithm is a sequential variant of a local search algorithm. We show by experiments that the proposed approach provides better overall optimization of both mean squared error and cluster balance compared to existing methods.","PeriodicalId":414924,"journal":{"name":"Applied Computing and Intelligence","volume":"65 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125708584","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}
Noah Gardner, John Paul Hellenbrand, Anthony Phan, Haige Zhu, Z. Long, Min Wang, C. Penick, Chih-Cheng Hung
Ant cuticle texture presumably provides some type of function, and therefore is useful to research for ecological applications and bioinspired designs. In this study, we employ statistical image texture analysis and deep machine learning methods to classify similar ant species based on morphological features. We establish a public database of ant cuticle images for research. We provide a comparative study of the performance of image texture classification and deep machine learning methods on this ant cuticle dataset. Our results show that the deep learning methods give higher accuracy than statistical methods in recognizing ant cuticle textures. Our experiments also reveal that the deep learning networks designed for image texture performs better than the general deep learning networks.
{"title":"Investigation of ant cuticle dataset using image texture analysis","authors":"Noah Gardner, John Paul Hellenbrand, Anthony Phan, Haige Zhu, Z. Long, Min Wang, C. Penick, Chih-Cheng Hung","doi":"10.3934/aci.2022008","DOIUrl":"https://doi.org/10.3934/aci.2022008","url":null,"abstract":"Ant cuticle texture presumably provides some type of function, and therefore is useful to research for ecological applications and bioinspired designs. In this study, we employ statistical image texture analysis and deep machine learning methods to classify similar ant species based on morphological features. We establish a public database of ant cuticle images for research. We provide a comparative study of the performance of image texture classification and deep machine learning methods on this ant cuticle dataset. Our results show that the deep learning methods give higher accuracy than statistical methods in recognizing ant cuticle textures. Our experiments also reveal that the deep learning networks designed for image texture performs better than the general deep learning networks.","PeriodicalId":414924,"journal":{"name":"Applied Computing and Intelligence","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125935630","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}
Deep learning methods may decline in their performance when the number of labelled training samples is limited, in a scenario known as few-shot learning. The methods may even degrade the accuracy in classifying instances of classes that have not been seen previously, called zero-shot learning. While the classification results achieved by the zero-shot learning methods are steadily improved, different problem settings, and diverse experimental setups have emerged. It becomes difficult to measure fairly the effectiveness of each proposed method, thus hindering further research in the field. In this article, a comprehensive survey is given on the methodology, implementation, and fair evaluations for practical and applied computing facets on the recent progress of zero-shot learning.
{"title":"A comprehensive survey of zero-shot image classification: methods, implementation, and fair evaluation","authors":"Guanyu Yang, Zihan Ye, Rui Zhang, Kaizhu Huang","doi":"10.3934/aci.2022001","DOIUrl":"https://doi.org/10.3934/aci.2022001","url":null,"abstract":"Deep learning methods may decline in their performance when the number of labelled training samples is limited, in a scenario known as few-shot learning. The methods may even degrade the accuracy in classifying instances of classes that have not been seen previously, called zero-shot learning. While the classification results achieved by the zero-shot learning methods are steadily improved, different problem settings, and diverse experimental setups have emerged. It becomes difficult to measure fairly the effectiveness of each proposed method, thus hindering further research in the field. In this article, a comprehensive survey is given on the methodology, implementation, and fair evaluations for practical and applied computing facets on the recent progress of zero-shot learning.","PeriodicalId":414924,"journal":{"name":"Applied Computing and Intelligence","volume":"518 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123298164","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}
In this paper, optimal control of an inverted pendulum on a cart system is studied. Since the nonlinear structure of the system is complex, and in the presence of random disturbances, optimization and control of the motion of the system become more challenging. For handling this system, a discrete-time stochastic optimal control problem for the system is described, where the external force is considered as the control input. By defining a loss function, namely, the mean squared errors to be minimized, the stochastic approximation (SA) approach is applied to estimate the state dynamics. In addition, the Hamiltonian function is defined, and the first-order necessary conditions are derived. The gradient of the cost function is determined so that the SA approach is employed to update the control sequences. For illustration, considering the values of the related parameters in the system, the discrete-time stochastic optimal control problem is solved iteratively by using the SA algorithm. The simulation results show that the state estimation and the optimal control law design are well performed with the SA algorithm, and the motion of the inverted pendulum cart is addressed satisfactorily. In conclusion, the efficiency of the SA approach for solving the inverted pendulum on a cart system is verified.
{"title":"State estimation and optimal control of an inverted pendulum on a cart system with stochastic approximation approach","authors":"Xian Wen Sim, S. Kek, Sy Yi Sim","doi":"10.3934/aci.2023005","DOIUrl":"https://doi.org/10.3934/aci.2023005","url":null,"abstract":"\u0000 In this paper, optimal control of an inverted pendulum on a cart system is studied. Since the nonlinear structure of the system is complex, and in the presence of random disturbances, optimization and control of the motion of the system become more challenging. For handling this system, a discrete-time stochastic optimal control problem for the system is described, where the external force is considered as the control input. By defining a loss function, namely, the mean squared errors to be minimized, the stochastic approximation (SA) approach is applied to estimate the state dynamics. In addition, the Hamiltonian function is defined, and the first-order necessary conditions are derived. The gradient of the cost function is determined so that the SA approach is employed to update the control sequences. For illustration, considering the values of the related parameters in the system, the discrete-time stochastic optimal control problem is solved iteratively by using the SA algorithm. The simulation results show that the state estimation and the optimal control law design are well performed with the SA algorithm, and the motion of the inverted pendulum cart is addressed satisfactorily. In conclusion, the efficiency of the SA approach for solving the inverted pendulum on a cart system is verified.\u0000","PeriodicalId":414924,"journal":{"name":"Applied Computing and Intelligence","volume":"83 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130924268","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}
When planning the development of future energy resources, electrical infrastructure, transportation networks, agriculture, and many other societally significant systems, policy makers require accurate and high-resolution data reflecting different climate scenarios. There is widely documented evidence that perceptual loss can be used to generate perceptually realistic results when mapping low-resolution inputs to high-resolution outputs, but its application is limited to images at present. In this paper, we study the perceptual loss when increasing the resolution of raw precipitation data by ×4 and ×8 under training modes of CNN and GAN. We examine the difference in the perceptual loss calculated by using different layers of feature maps and demonstrate how low- and mid-level feature maps can yield comparable results to pixel-wise loss. In particular, from both qualitative and quantitative points of view, Conv2_1 and Conv3_1 are the best compromises between obtaining detailed information and maintaining the overall error in our case. We propose a new approach to benefit from perceptual loss while considering the characteristics of climate data. We show that in comparison to calculating perceptual loss directly for the entire sample, our proposed approach can obtain detailed information of extreme events regions while reducing error.
{"title":"Perceptual loss function for generating high-resolution climate data","authors":"Yang Wang, H. Karimi","doi":"10.3934/aci.2022009","DOIUrl":"https://doi.org/10.3934/aci.2022009","url":null,"abstract":"\u0000 When planning the development of future energy resources, electrical infrastructure, transportation networks, agriculture, and many other societally significant systems, policy makers require accurate and high-resolution data reflecting different climate scenarios. There is widely documented evidence that perceptual loss can be used to generate perceptually realistic results when mapping low-resolution inputs to high-resolution outputs, but its application is limited to images at present. In this paper, we study the perceptual loss when increasing the resolution of raw precipitation data by ×4 and ×8 under training modes of CNN and GAN. We examine the difference in the perceptual loss calculated by using different layers of feature maps and demonstrate how low- and mid-level feature maps can yield comparable results to pixel-wise loss. In particular, from both qualitative and quantitative points of view, Conv2_1 and Conv3_1 are the best compromises between obtaining detailed information and maintaining the overall error in our case. We propose a new approach to benefit from perceptual loss while considering the characteristics of climate data. We show that in comparison to calculating perceptual loss directly for the entire sample, our proposed approach can obtain detailed information of extreme events regions while reducing error.\u0000","PeriodicalId":414924,"journal":{"name":"Applied Computing and Intelligence","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128197874","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}
In obesity studies, several researchers have been applying machine learning tools to identify factors affecting human body weight. However, a proper review of strength, limitations and evaluation metrics of machine learning algorithms in obesity is lacking. This study reviews the status of application of machine learning algorithms in obesity studies and to identify strength and weaknesses of these methods. A scoping review of paper focusing on obesity was conducted. PubMed and Scopus databases were searched for the application of machine learning in obesity using different keywords. Only English papers in adult obesity between 2014 and 2019 were included. Also, only papers that focused on controllable factors (e.g., nutrition intake, dietary pattern and/or physical activity) were reviewed in depth. Papers on genetic or childhood obesity were excluded. Twenty reviewed papers used machine learning algorithms to identify the relationship between the contributing factors and obesity. Regression algorithms were widely applied. Other algorithms such as neural network, random forest and deep learning were less exploited. Limitations regarding data priori assumptions, overfitting and hyperparameter optimization were discussed. Performance metrics and validation techniques were identified. Machine learning applications are positively impacting obesity research. The nature and objective of a study and available data are key factors to consider in selecting the appropriate algorithms. The future research direction is to further explore and take advantage of the modern methods, i.e., neural network and deep learning, in obesity studies.
{"title":"A review of the application of machine learning in adult obesity studies","authors":"M. Alkhalaf, Ping Yu, Jun Shen, Chao Deng","doi":"10.3934/aci.2022002","DOIUrl":"https://doi.org/10.3934/aci.2022002","url":null,"abstract":"\u0000 In obesity studies, several researchers have been applying machine learning tools to identify factors affecting human body weight. However, a proper review of strength, limitations and evaluation metrics of machine learning algorithms in obesity is lacking. This study reviews the status of application of machine learning algorithms in obesity studies and to identify strength and weaknesses of these methods. A scoping review of paper focusing on obesity was conducted. PubMed and Scopus databases were searched for the application of machine learning in obesity using different keywords. Only English papers in adult obesity between 2014 and 2019 were included. Also, only papers that focused on controllable factors (e.g., nutrition intake, dietary pattern and/or physical activity) were reviewed in depth. Papers on genetic or childhood obesity were excluded. Twenty reviewed papers used machine learning algorithms to identify the relationship between the contributing factors and obesity. Regression algorithms were widely applied. Other algorithms such as neural network, random forest and deep learning were less exploited. Limitations regarding data priori assumptions, overfitting and hyperparameter optimization were discussed. Performance metrics and validation techniques were identified. Machine learning applications are positively impacting obesity research. The nature and objective of a study and available data are key factors to consider in selecting the appropriate algorithms. The future research direction is to further explore and take advantage of the modern methods, i.e., neural network and deep learning, in obesity studies.\u0000","PeriodicalId":414924,"journal":{"name":"Applied Computing and Intelligence","volume":"64 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114683803","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}
This paper presents a new class of games: location puzzle games. It combines puzzle games with the use of the geographical location. The game class is closely related to location-based games except that no physical movement in the real world is needed as in most mobile location-based games. For example, we present a game called Puzzle-Mopsi, which asks users to match a given set of images with the locations shown on the map. In addition to local knowledge, the game requires logical skills as the number of possible matches grows exponentially with the number of images. Small-scale experiments show that the players found the game interesting and that the difficulty increases with the number of targets and decreases with the player's familiarity with the area.
{"title":"Puzzle-Mopsi: a location-puzzle game","authors":"P. Fränti, Lingyi Kong","doi":"10.3934/aci.2023001","DOIUrl":"https://doi.org/10.3934/aci.2023001","url":null,"abstract":"\u0000\u0000This paper presents a new class of games: location puzzle games. It combines puzzle games with the use of the geographical location. The game class is closely related to location-based games except that no physical movement in the real world is needed as in most mobile location-based games. For example, we present a game called Puzzle-Mopsi, which asks users to match a given set of images with the locations shown on the map. In addition to local knowledge, the game requires logical skills as the number of possible matches grows exponentially with the number of images. Small-scale experiments show that the players found the game interesting and that the difficulty increases with the number of targets and decreases with the player's familiarity with the area.\u0000\u0000","PeriodicalId":414924,"journal":{"name":"Applied Computing and Intelligence","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116100100","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}
Evolutionary algorithm is one of the optimization techniques. Cat swarm optimization (CSO)-based algorithm is frequently used in many applications for solving challenging optimization problems. In this paper, the tracing mode in CSO is modified to reduce the number of user-defined parameters and weaken the sensitivity to the parameter values. In addition, a mode ratio control scheme for switching individuals between different movement modes and a search boosting strategy are proposed. The obtained results from our method are compared with the modified CSO without the proposed strategy, the original CSO, the particle swarm optimization (PSO) and differential evolution (DE) with three commonly-used DE search schemes. Six test functions from IEEE congress on evolutionary competition (CEC) are used to evaluate the proposed methods. The overall performance is evaluated by the average ranking over all test results. The ranking result indicates that our proposed method outperforms the other methods compared.
{"title":"Adjustable mode ratio and focus boost search strategy for cat swarm optimization","authors":"Pei-wei Tsai, Xingsi Xue, Jing Zhang, V. Istanda","doi":"10.3934/aci.2021005","DOIUrl":"https://doi.org/10.3934/aci.2021005","url":null,"abstract":"Evolutionary algorithm is one of the optimization techniques. Cat swarm optimization (CSO)-based algorithm is frequently used in many applications for solving challenging optimization problems. In this paper, the tracing mode in CSO is modified to reduce the number of user-defined parameters and weaken the sensitivity to the parameter values. In addition, a mode ratio control scheme for switching individuals between different movement modes and a search boosting strategy are proposed. The obtained results from our method are compared with the modified CSO without the proposed strategy, the original CSO, the particle swarm optimization (PSO) and differential evolution (DE) with three commonly-used DE search schemes. Six test functions from IEEE congress on evolutionary competition (CEC) are used to evaluate the proposed methods. The overall performance is evaluated by the average ranking over all test results. The ranking result indicates that our proposed method outperforms the other methods compared.","PeriodicalId":414924,"journal":{"name":"Applied Computing and Intelligence","volume":"24 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125051057","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：