Proceedings of the 3rd International Conference on Advances in Artificial Intelligence最新文献

{"title":"Deep Learning-based Detection of Overlapping Cells","authors":"A. Paulauskaitė-Tarasevičienė, K. Sutiene, Justas Valotka, V. Raudonis, T. Iešmantas","doi":"10.1145/3369114.3369120","DOIUrl":"https://doi.org/10.1145/3369114.3369120","url":null,"abstract":"Detection and quantification of cells in general is one of the key challenges in many clinical trials for disease diagnosis and monitoring. Automation of this task enables quantitative analysis of digital images with a high processing rate, which is a support to pathologist at various kind of analyses. Recent studies have already indicated that deep learning usually yield superior accuracy in the field of digital pathology. One of the challenges tackled by the researches is to detect cells in images when cells are highly overlapped, over illuminated or partially occluded with the noise. Therefore, we focused on two conceptually different deep learning models, specifically U-Net and Mask R-CNN, in order to evaluate their capability and performance on the detection of overlapping cells. The dataset used in the study contains different types of images, possible observed under different lighting conditions, and the amount of target cells may range from tens to thousands, therefore the algorithm is required to be flexible enough.","PeriodicalId":262468,"journal":{"name":"Proceedings of the 3rd International Conference on Advances in Artificial Intelligence","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126526686","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}

{"title":"Modelling Genomics Datasets to Optimize Its Usage Patterns for Efficient Analytics","authors":"Z. A. Shah, A. Anjum, M. Abouelhoda","doi":"10.1145/3369114.3369148","DOIUrl":"https://doi.org/10.1145/3369114.3369148","url":null,"abstract":"Data intensive computing requires fast access to the input data. Currently in an HPC environment, the computation part is managed by the job scheduler, whereas data access is unmanaged and reply completely on the underlying file system. The job scheduler tracks system parameters such as RAM, CPU and utilize this for efficient computation. However, it is the responsibility of the developer to take care of the data related attributes and their traffic over the network. To solve this issue, we suggest data models for genomics data which can be used to guide the computation engine from data management perspective in addition to the usual system and computation parameters -- We will demonstrated that enabling the job scheduler with info about the data is an important steps towards efficient fully automated data analysis.","PeriodicalId":262468,"journal":{"name":"Proceedings of the 3rd International Conference on Advances in Artificial Intelligence","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116347138","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}

{"title":"The use of Local Binary Pattern (LBP) feature extraction Members of the mud crab genus Scylla","authors":"Rozniza Ali, M. Fauzi","doi":"10.1145/3369114.3369117","DOIUrl":"https://doi.org/10.1145/3369114.3369117","url":null,"abstract":"In the member of mud crab, species of Scylla are the most traded seafood commodity in Asia and the culturing practice is already done in most of Asia country a few years ago. The demand for mud crabs has increased rapidly over the last decade, providing great potential for the development of the mud crab aquaculture industry. But, there is still unsolved problem where limitation of knowledge in identification is limited due to similar colours and feature characteristic. Thus, this study proposed an automatic technique that can evaluate and produce the subset of mud crab genus Scylla features by using Local Binary Pattern (LBP) as a feature extraction tool. The main objective of the study is to find the optimal subset of mud crab genus Scylla features from a carapace images dataset. Based on 153 extracted features chosen by LBP features selection methods, the accuracy rates of three classification algorithms were obtained for analysis. The results from the MatLab experiment demonstrated that, the LBP method produced an accuracy under 60% for entire classifier.","PeriodicalId":262468,"journal":{"name":"Proceedings of the 3rd International Conference on Advances in Artificial Intelligence","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131094158","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}

{"title":"Probability κ-means Clustering for Neural Network Architecture","authors":"Ka‐Hou Chan, W. Ke, S. Im","doi":"10.1145/3369114.3369147","DOIUrl":"https://doi.org/10.1145/3369114.3369147","url":null,"abstract":"Cluster analysis indices are aimed at classifying the elements used to estimate the quality of the categories based on their similarity. It is a challenging task because, with the same data set, there may be many partitions that fit natural groupings of a given data set. Applications of clustering can include pattern recognition, image analysis and information retrieval. We propose an approach based on the κ-means concept that clustering centers more often have a higher density than their neighbors: then we use a probability κ-means algorithm to achieve fuzzy clustering in continuous form over a relatively large distance from other points with higher densities. Further, in order to follow the mainstream neural network architecture, we define a favorable activation function and corresponding loss function for the clustering iteration. Our method come from the basis of a clustering procedure in which the number of clusters arises intuitively and clusters are achieved regardless of the high dimensions. We demonstrate the result of complete algorithm on several clustering test cases.","PeriodicalId":262468,"journal":{"name":"Proceedings of the 3rd International Conference on Advances in Artificial Intelligence","volume":"30 1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131390852","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}

{"title":"Capturing Smoker in Non-Smoking Room","authors":"D. H. Hareva, Samuel Teguh","doi":"10.1145/3369114.3369159","DOIUrl":"https://doi.org/10.1145/3369114.3369159","url":null,"abstract":"In developing countries legal actions regarding violations of smokers in non-smoking areas are still weak. This will be very disturbing especially those who care for their own health. Eastern cultures are generally reluctant to treat smokers who violate the rules directly. Therefore, other methods are needed, such as creating a set of cigarette smoke detection systems that can warn smokers not to do it in a place that is forbidden to smoke. This device is expected to detect cigarette smoke, then at the same time take photos of suspected smokers as evidence of violations. Warnings will be carried out by authorized persons, if the alarm signal of cigarette smoke is not also heeded. The device consists of NodeMCU, MQ-2, piezo speakers, and ArduCAM OV2640. When the device detects cigarette smoke, it will push a beep sound, capture images of the room, and immediately send images to the Cloud database. A few moments later the mobile phone supervisor of the room will get a proof of the incident and immediately take an action to reprimand and arrest the perpetrators. The testing of cigarette smoke detection devices was carried out in a cardboard room with a volume of 50cm x 50cm x 75cm. Detection was carried out at 4 different points from the source of cigarette smoke each of 6 x experiments or 24 x experiments. It was found, that the speed of detection of cigarette smoke was achieved on average 14.4 seconds with the device located above the source of cigarette smoke (75cm).","PeriodicalId":262468,"journal":{"name":"Proceedings of the 3rd International Conference on Advances in Artificial Intelligence","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131498113","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}

{"title":"Improving Metrics to Rank the Research Performance","authors":"Yaniasih Yaniasih, Aris Yaman, Ambar Yoganingrum, F. Saputra, H. Susanto, S. Riyanto","doi":"10.1145/3369114.3369317","DOIUrl":"https://doi.org/10.1145/3369114.3369317","url":null,"abstract":"Citations are increasingly used to consider research performance. Subsequently, the mathematical and statistical approach by developing metrics are applied to ease the performance measurement. However, almost all the existing metrics only calculate the quantity, not the quality of the citation. This approach is considered unfair. Hence, this study aims to explore the other variables influencing the quality of the research performance ranking. This paper conducted a literature survey finding the citation patterns, which is assumed fairer. Next, the patterns of the citation time window, in-text citation, and citation exclusivity to improve the metrics are suggested. This paper discussed the opportunities and the challenges in combining those patterns in one metric, especially for measuring the research performance in Indonesia.","PeriodicalId":262468,"journal":{"name":"Proceedings of the 3rd International Conference on Advances in Artificial Intelligence","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124069873","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}

{"title":"Improving the Convergence Speed of Deep Neural Networks with Biased Sampling","authors":"G. Ioannou, Thanos Tagaris, A. Stafylopatis","doi":"10.1145/3369114.3369116","DOIUrl":"https://doi.org/10.1145/3369114.3369116","url":null,"abstract":"Deep Neural Networks (DNNs) are usually trained using a stochastic gradient descent algorithm or its variants. This involves the gradient being computed on small batches of samples. The decision of which samples will be selected is an important step in the training process, as it can greatly affect the DNN's performance. This paper proposes a novel method for batch selection, that involves replacing easy training examples with difficult ones every epoch, which results in an acceleration of the training process of deep networks. The resulting algorithm was tested on four datasets, representing both regression (Boston Housing) and image classification (MNIST, CIFAR10, CIFAR100) problems. The proposed approach, called Batch Selection with Biased Sampling (BSBS), yields faster convergence rates. Experiments indicate that the model can be trained in up to 50% less epochs.","PeriodicalId":262468,"journal":{"name":"Proceedings of the 3rd International Conference on Advances in Artificial Intelligence","volume":"127 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117348788","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}

{"title":"An AI Driven Real-time 3-D Representation of an Off-shore WT for Fault Diagnosis and Monitoring","authors":"Amin Amini, J. Kanfoud, T. Gan","doi":"10.1145/3369114.3369141","DOIUrl":"https://doi.org/10.1145/3369114.3369141","url":null,"abstract":"In recent years, digital twin (DT) has been an efficient and reliable representation tool for Industry 4.0 in general and fault diagnosis and monitoring, especially in machinery consisting of several components working in tandem that introduce a large amount of data to be processed. Having a cloud-based and platform-independent system that can interpret Big Data in real-time and represent the processed information in a user-friendly manner offers the opportunity not only for improving fault diagnosis and troubleshooting automation, but also offering a better representation of data for operators, technicians and engineers. This not only decreases the repair costs by enabling constant components' state of health awareness, but also increases the efficacy of the apparatus by lowering the downtimes due to faults and issues. In this paper (WindTwin), we propose a real-time, interactive and dynamic 3-D neural-network model that provides important information of a small-scale wind turbine (WT) machinery setup including the state of health of different components (bearings, shafts) based on monitoring sensors (vibrations and temperature) and operations conditions (rotation speed). The system features real-time data processing of six sensors with eight channels including two single axis-accelerometer, one tri-axial accelerometer, one acoustics (microphone), one temperature and one light sensor (for rotational speed) with sampling data collection rate of 51.2 kHz achieved through an edge developed machine learning algorithm based on Artificial Neural Network (ANN) algorithm using single hidden-layer feed forward neural network (SLFN) and gradient-based backpropagation (BP) training algorithm. The system also provides platform-independent (desktop and mobile companion app) user interfaces thanks to its cloud-based data processing integration. The prototype system proved to have on average 83.33 % accuracy for the vibration-based condition monitoring model across the three different rotation speed of 15Hz, 9Hz and 12Hz with 1 second latency for different fault conditions to be reflected on the DT counterpart.","PeriodicalId":262468,"journal":{"name":"Proceedings of the 3rd International Conference on Advances in Artificial Intelligence","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126953103","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}

{"title":"Ultra-Scalable Simulations of Networks of Polarized Evolutionary Processors","authors":"Sandra Gómez Canaval, V. Mitrana, M. Păun, Stanislav Vararuk","doi":"10.1145/3369114.3369156","DOIUrl":"https://doi.org/10.1145/3369114.3369156","url":null,"abstract":"The Network of Polarized Evolutionary Processors (NPEP) is a rather new variant of the bio-inspired computing model called Network of Evolutionary Processors (NEP). This model, together with its variants, is able to provide theoretical feasible solutions to hard computational problems. NPEPE is a software engine able to simulate NPEP which is deployed over Giraph, an ultra-scalable platform based on the Bulk Synchronous Parallel (BSP) programming model. Rather surprisingly, the BSP model and the underlying architecture of NPEP have many common points. Moreover, these similarities are also shared with all variants in the NEP family. We take advantage of these similarities and propose an extension of NPEPE (named gNEP) that enhances it to simulate any variant of the NEP's family. Our extended gNEP framework, presents a twofold contribution. Firstly, a flexible architecture able to extend software components in order to include other NEP models (including the seminal NEP model and new ones). Secondly, a component able to translate input configuration files representing the instance of a problem and an algorithm based on different variants of the NEP model into some suitable input files for gNEP framework. In this work, we simulate a solution to the \"3-colorability\" problem which is based on NPEP. We compare the results for a specific experiment using NPEPE engine and gNEP. Moreover, we show several experiments in the aim of studying, in a preliminary way, the scalability offered by gNEP to easily deploy and execute instances of problems requiring more intensive computations.","PeriodicalId":262468,"journal":{"name":"Proceedings of the 3rd International Conference on Advances in Artificial Intelligence","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123055341","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}

{"title":"Comparing SVM and GLM in Calculating Insurance Premium for Flight Delay","authors":"Samuel Lukas, Dina Stefani, Petrus Widjaja","doi":"10.1145/3369114.3369160","DOIUrl":"https://doi.org/10.1145/3369114.3369160","url":null,"abstract":"The development of land transportation in Indonesia does not discourage users to use flight transportation modes. This is due to the shorter travel time and the high safety factor. Sometimes flight delay and cancelation cannot be avoided, which cause the airline's company must give compensation to customers. However, customers feel dissatisfied because compensation is considered inadequate even though it has followed government regulations. Airlines, on the other hand, also feel reluctant to improve their services, one of the possible reasons is the cost of customer compensation is far cheaper than the cost of improving services. The purpose of this study is to propose a premium calculation to overcome this problem to satisfy both parties. If the premium is purchased by users whose price is included in the ticket price, the airline can receive additional funds to improve their services. There are several premium calculation techniques, this paper discusses three calculation models. They are standard deviation, Generalize Linier Model, and Support Vector machine. The result shows that the premium of the three methods for an airline departs from a certain city is slightly different. However, SVM gives the better result than that of others.","PeriodicalId":262468,"journal":{"name":"Proceedings of the 3rd International Conference on Advances in Artificial Intelligence","volume":"61 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127095624","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}