Towards a real-time processing framework based on improved distributed recurrent neural network variants with fastText for social big data analytics
به سمت یک چارچوب پردازش در زمان واقعی بر اساس بهبود انواع شبکه عصبی مکرر توزیع شده با fastText برای تجزیه و تحلیل داده های بزرگ اجتماعی-2020
Big data generated by social media stands for a valuable source of information, which offers an excellent opportunity to mine valuable insights. Particularly, User-generated contents such as reviews, recommendations, and users’ behavior data are useful for supporting several marketing activities of many companies. Knowing what users are saying about the products they bought or the services they used through reviews in social media represents a key factor for making decisions. Sentiment analysis is one of the fundamental tasks in Natural Language Processing. Although deep learning for sentiment analysis has achieved great success and allowed several firms to analyze and extract relevant information from their textual data, but as the volume of data grows, a model that runs in a traditional environment cannot be effective, which implies the importance of efficient distributed deep learning models for social Big Data analytics. Besides, it is known that social media analysis is a complex process, which involves a set of complex tasks. Therefore, it is important to address the challenges and issues of social big data analytics and enhance the performance of deep learning techniques in terms of classification accuracy to obtain better decisions. In this paper, we propose an approach for sentiment analysis, which is devoted to adopting fastText with Recurrent neural network variants to represent textual data efficiently. Then, it employs the new representations to perform the classification task. Its main objective is to enhance the performance of well-known Recurrent Neural Network (RNN) variants in terms of classification accuracy and handle large scale data. In addition, we propose a distributed intelligent system for real-time social big data analytics. It is designed to ingest, store, process, index, and visualize the huge amount of information in real-time. The proposed system adopts distributed machine learning with our proposed method for enhancing decision-making processes. Extensive experiments conducted on two benchmark data sets demonstrate that our proposal for sentiment analysis outperforms well-known distributed recurrent neural network variants (i.e., Long Short-Term Memory (LSTM), Bidirectional Long Short-Term Memory (BiLSTM), and Gated Recurrent Unit (GRU)). Specifically, we tested the efficiency of our approach using the three different deep learning models. The results show that our proposed approach is able to enhance the performance of the three models. The current work can provide several benefits for researchers and practitioners who want to collect, handle, analyze and visualize several sources of information in real-time. Also, it can contribute to a better understanding of public opinion and user behaviors using our proposed system with the improved variants of the most powerful distributed deep learning and machine learning algorithms. Furthermore, it is able to increase the classification accuracy of several existing works based on RNN models for sentiment analysis.
Keywords: Big data | FastText | Recurrent neural networks | LSTM | BiLSTM | GRU | Natural language processing | Sentiment analysis | Social big data analytics
Forecasting across time series databases using recurrent neural networks on groups of similar series: A clustering approach
پیش بینی پیش بینی پایگاه داده های سری زمانی با استفاده از شبکه های عصبی مکرر در گروه های مشابه سری: یک روش خوشه بندی-2020
With the advent of Big Data, nowadays in many applications databases containing large quantities of sim- ilar time series are available. Forecasting time series in these domains with traditional univariate fore- casting procedures leaves great potentials for producing accurate forecasts untapped. Recurrent neural networks (RNNs), and in particular Long Short Term Memory (LSTM) networks, have proven recently that they are able to outperform state-of-the-art univariate time series forecasting methods in this context, when trained across all available time series. However, if the time series database is heterogeneous, ac- curacy may degenerate, so that on the way towards fully automatic forecasting methods in this space, a notion of similarity between the time series needs to be built into the methods. To this end, we present a prediction model that can be used with different types of RNN models on subgroups of similar time series, which are identified by time series clustering techniques. We assess our proposed methodology using LSTM networks, a widely popular RNN variant, together with various clustering algorithms, such as kMeans, DBScan, Partition Around Medoids (PAM), and Snob. Our method achieves competitive results on benchmarking datasets under competition evaluation procedures. In particular, in terms of mean sMAPE accuracy it consistently outperforms the baseline LSTM model, and outperforms all other methods on the CIF2016 forecasting competition dataset.
Keywords: Big data forecasting | RNN | LSTM | Time series clustering | Neural networks
روش یادگیری متخاصم عمیق و چند مرحله ای ، برای باز شناسی شخص مبتنی بر ویدئو
سال انتشار: 2020 - تعداد صفحات فایل pdf انگلیسی: 13 - تعداد صفحات فایل doc فارسی: 42
بازشناسی شخص (re-ID) بر مبنای ویدئو را میتوان به عنوان فرآیند تطبیق تصویر یک فرد از طریق دیدهای مختلف دوربین که به وسیله ی تصاویر ویدئویی ناهم راستا گرفته شده است، در نظر گرفت. روش هایی که برای اینکار وجود دارند، از سیگنال های نظارتی برای بهینه سازی فضای پیش روی دوربین استفاده نموده که تحت این شرایط، فاصله ی بین ویدئوها بیشینه سازی/کمینه سازی میشود. البته این کار باعث شده تا برچسب گذاری افراد در سطح دید های ویدئو بسیار زیاد شده و باعث شده تا نتوان آنها را به خوبی بر روی دوربین های شبکه بندی شده ی بزرگ مقیاس بندی کرد. همچنین خاطر نشان شده است که یادگیری نمایش های مختلف ویدئویی و آنهم به وسیله ی عدم تغییر دید دوربین را نمیتوان انجام داد چرا که ویژگی های تصویر، هر کدام دارای توزیع های مختلف مختص به خود میباشند. بنابراین تطبیق ویدئوها برای باز شناسی افراد، نیاز به مدل هایی انعطاف پذیر برای بدست آوردن پویایی های موجود در مشاهدات ویدئویی و یادگیری دیدهای ثابت از طریق دسترسی به نمونه های آموزشی برچسب دار و محدود دارد. در این مقاله قصد داریم یک روش مبتنی بر یادگیری عمیق چند مرحله ای را برای باز شناسی یک فرد بر مبنای ویدئو ارائه دهیم و بتوانیم به یادگیری دیدهای قابل قیاسی از این فرد که متمایز هستند بپردازیم. روش پیشنهادی را بر روی شبکه های عصبی باز رخداد گر متغیر (VRNN) توسعه داده ایم و آنرا به منظور ایجاد متغیر های پنهان با وابستگی های موقت که بسیار متمایز بوده ولی در تطبیق تصاویر فرد از نظر دید ثابت میباشد، مورد یادگیری قرار داده ایم. آزمایش های وسیعی را بر روی سه مجموعه ی داده ای بنچ مارک انجام داده ایم و به صورت تجربی به اثبات قابلیت روش پیشنهادی مان در ایجاد ویژگی های موقتی و با یک دید ثابت و کارائی بالایی که به وسیله ی آن بدست آمده است خواهیم پرداخت.
کلمات کلیدی: باز شناسی شخص مبتنی بر ویدئو | شبکه های عصبی باز رخدادگر متغیر | یادگیری متخاصم
|مقاله ترجمه شده|
Static malware detection and attribution in android byte-code through an end-to-end deep system
شناسایی بدافزارهای استاتیکی و انتساب در بایت کد اندرویدی از طریق یک سیستم عمیق انتها به انتها-2020
Android reflects a revolution in handhelds and mobile devices. It is a virtual machine based, an open source mobile platform that powers millions of smartphone and devices and even a larger no. of applications in its ecosystem. Surprisingly in a short lifespan, Android has also seen a colossal expansion in application malware with 99% of the total malware for smartphones being found in the Android ecosystem. Subsequently, quite a few techniques have been proposed in the literature for the analysis and detection of these malicious applications for the Android platform. The increasing and diversified nature of Android malware has immensely attenuated the usefulness of prevailing malware detectors, which leaves Android users susceptible to novel malware. Here in this paper, as a remedy to this problem, we propose an anti-malware system that uses customized learning models, which are sufficiently deep, and are ’End to End deep learning architectures which detect and attribute the Android malware via opcodes extracted from application bytecode’. Our results show that Bidirectional long short-term memory (BiLSTMs) neural networks can be used to detect static behavior of Android malware beating the state-of-the-art models without using handcrafted features. For our experiments in our system, we also choose to work with distinct and independent deep learning models leveraging sequence specialists like recurrent neural networks, Long Short Term Memory networks and its Bidirectional variation as well as those are more usual neural architectures like a network of all connected layers(fully connected), deep convnets, Diabolo network (autoencoders) and generative graphical models like deep belief networks for static malware analysis on Android. To test our system, we have also augmented a bytecode dataset from three open and independently maintained state-of-the-art datasets. Our bytecode dataset, which is on an order of magnitude large, essentially suffice for our experiments. Our results suggests that our proposed system can lead to better design of malware detectors as we report an accuracy of 0.999 and an F1-score of 0.996 on a large dataset of more than 1.8 million Android applications.
Keywords: End-to-end architecture | Malware analysis | Deep neural networks | Android and big data
A new hybrid deep learning model for human action recognition
یک مدل جدید یادگیری عمیق ترکیبی برای شناخت عملکرد انسان-2019
Human behavior has been always an important factor in social communication. The human activity and action recognition are all clues that facilitate the analysis of human behavior. Human action recognition is an important challenge in a variety of application including human-computer interaction and intelligent video surveillance to enhance security in different domains. The evaluation algorithm relies on the proper extraction and the learning data. The success of the deep learning led to many imposing results in several contexts that include neural network. Here the emergence of Gated Recurrent Neural Networks with increased computation powers is being adopted for sequential data and video classification. However, to have an efficient classifier for assigning the class label, it is very necessary to have a strong features vector. Features are the most important information in each data. Indeed, features extraction can influence on the performance of the algorithm and the computation complexity. This paper proposes a novel approach for human action recognition based on hybrid deep learning model. The proposed approach is evaluated on the challenging UCF Sports, UCF101 and KTH datasets. An average of 96.3% is obtained when we have tested on KTH dataset
Keywords: Deep learning | Recurrent Neural Networks | Gated Recurrent Unit | Video classification | Motion detection
Slanderous user detection with modified recurrent neural networks in recommender system
تشخیص کاربر مخدوش با شبکه های عصبی بازرخدادگر اصلاح شده در سیستم توصیه گر-2019
We focus on how to tackle a unique multi-view unsupervised issue: slanderous user de- tection, with recurrent neural networks to benefit recommender systems. In real-world recommender systems, some consumers always give fake reviews and low ratings to the items they bought on purpose. In order to ensure their profits, these slanderous users make a semantic gap between their ratings and reviews to avoid detection, which makes slanderous user detection a more difficult problem. On some occasions, they give a false low rating with a positive review which confuse recommender systems, and vice versa. To address the above problem, in this paper, we propose a novel recommendation frame- work: Slanderous user Detection Recommender System (SDRS). In SDRS, we design a Hier- archical Dual-Attention recurrent Neural network (HDAN) with a modified GRU (mGRU) to compute an opinion level for reviews. Then a joint filtering method is proposed to catch the gap between ratings and reviews. With joint filtering, slanderous users can be de- tected and omitted. Finally, a modified non-negative matrix factorization is proposed to make recommendations. Extensive experiments are conducted in four datasets: Amazon, Yelp, Taobao, and Jingdong, in which the results demonstrate that our proposed method can detect slanderous users and make accurate recommendations in a uniform framework. Also, with slanderous user detection, some state-of-the-art recommendation systems can be benefited.
Keywords: Slanderous user detection | Recommender systems | Recurrent neural networks
Automating orthogonal defect classification using machine learning algorithms
خودکارسازی طبقه بندی نقص متعامد با استفاده از الگوریتم های یادگیری ماشین-2019
Software systems are increasingly being used in business or mission critical scenarios, where the presence of certain types of software defects, i.e., bugs, may result in catastrophic consequences (e.g., financial losses or even the loss of human lives). To deploy systems in which we can rely on, it is vital to understand the types of defects that tend to affect such systems. This allows developers to take proper action, such as adapting the development process or redirecting testing efforts (e.g., using a certain set of testing techniques, or focusing on certain parts of the system). Orthogonal Defect Classification (ODC) has emerged as a popular method for classifying software defects, but it requires one or more experts to categorize each defect in a quite complex and time-consuming process. In this paper, we evaluate the use of machine learning algorithms (k-Nearest Neighbors, Support Vector Machines, Naïve Bayes, Nearest Centroid, Random Forest and Recurrent Neural Networks) for automatic classification of software defects using ODC, based on unstructured textual bug reports. Experimental results reveal the difficulties in automatically classifying certain ODC attributes solely using reports, but also suggest that the overall classification accuracy may be improved in most of the cases, if larger datasets are used.
Index Terms : Software Defects | Bug Reports | Orthogonal Defect Classification | Machine Learning | Text Classification
Nemesyst: A hybrid parallelism deep learning-based framework applied for internet of things enabled food retailing refrigeration systems
Nemesyst: یک چارچوب مبتنی بر یادگیری عمیق موازی ترکیبی برای سیستم های تبرید خرده فروشی مواد غذایی توانا شده با اینترنت اشیا-2019
Deep learning has attracted considerable attention across multiple application domains, including com-puter vision, signal processing and natural language processing. Although quite a few single node deeplearning frameworks exist, such as tensorflow, pytorch and keras, we still lack a complete processingstructure that can accommodate large scale data processing, version control, and deployment, all whilestaying agnostic of any specific single node framework. To bridge this gap, this paper proposes a new,higher level framework, i.e. Nemesyst, which uses databases along with model sequentialisation to allowprocesses to be fed unique and transformed data at the point of need. This facilitates near real-time appli-cation and makes models available for further training or use at any node that has access to the databasesimultaneously. Nemesyst is well suited as an application framework for internet of things aggregatedcontrol systems, deploying deep learning techniques to optimise individual machines in massive net-works. To demonstrate this framework, we adopted a case study in a novel domain; deploying deeplearning to optimise the high speed control of electrical power consumed by a massive internet of thingsnetwork of retail refrigeration systems in proportion to load available on the UK National Grid (a demandside response). The case study demonstrated for the first time in such a setting how deep learning models,such as Recurrent Neural Networks (vanilla and Long-Short-Term Memory) and Generative AdversarialNetworks paired with Nemesyst, achieve compelling performance, whilst still being malleable to futureadjustments as both the data and requirements inevitably change over time.
Keywords:Deep learning | Databases | Distributed computing | Parallel computing | Demand side | responseRefrigeration | Internet of things
An empirical evaluation of deep learning for ICD-9 code assignment using MIMIC-III clinical notes
یک ارزیابی تجربی از یادگیری عمیق برای اختصاص کد ICD-9 با استفاده از یادداشتهای بالینی MIMIC-III-2019
Background and Objective: Code assignment is of paramount importance in many levels in modern hos- pitals, from ensuring accurate billing process to creating a valid record of patient care history. However, the coding process is tedious and subjective, and it requires medical coders with extensive training. This study aims to evaluate the performance of deep-learning-based systems to automatically map clinical notes to ICD-9 medical codes. Methods: The evaluations of this research are focused on end-to-end learning methods without manually defined rules. Traditional machine learning algorithms, as well as state-of-the-art deep learning methods such as Recurrent Neural Networks and Convolution Neural Networks, were applied to the Medical Infor- mation Mart for Intensive Care (MIMIC-III) dataset. An extensive number of experiments was applied to different settings of the tested algorithm. Results: Findings showed that the deep learning-based methods outperformed other conventional ma- chine learning methods. From our assessment, the best models could predict the top 10 ICD-9 codes with 0.6957 F 1 and 0.8967 accuracy and could estimate the top 10 ICD-9 categories with 0.7233 F 1 and 0.8588 accuracy. Our implementation also outperformed existing work under certain evaluation metrics. Conclusion: A set of standard metrics was utilized in assessing the performance of ICD-9 code assignment on MIMIC-III dataset. All the developed evaluation tools and resources are available online, which can be used as a baseline for further research
Keywords: Deep learning | Clinical notes | Machine learning | ICD-9 | Medical codes RNNs CNNs MIMIC-III Code assignment
Deep learning approaches for automatic detection of sleep apnea events from an electrocardiogram
رویکردهای یادگیری عمیق برای تشخیص خودکار رویدادهای sleep apnea از الکتروکاردیوگرام-2019
Background and Objective: This study demonstrates deep learning approaches with an aim to find the optimal method to automatically detect sleep apnea (SA) events from an electrocardiogram (ECG) signal. Methods: Six deep learning approaches were designed and implemented for automatic detection of SA events including deep neural network (DNN), one-dimensional (1D) convolutional neural networks (CNN), two-dimensional (2D) CNN, recurrent neural networks (RNN), long short-term memory, and gated- recurrent unit (GRU). Designed deep learning models were analyzed and compared in the performances. The ECG signal was pre-processed, normalized, and segmented into 10 s intervals. Subsequently, the sig- nal was converted into a 2D form for analysis in the 2D CNN model. A dataset collected from 86 patients with SA was used. The training set comprised data from 69 of the patients, while the test set contained data from the remaining 17 patients. Results: The accuracy of the best-performing model was 99.0%, and the 1D CNN and GRU models had 99.0% recall rates. Conclusions: The designed deep learning approaches performed better than those developed and tested in previous studies in terms of detecting SA events, and they could distinguish between apnea and hypopnea events using an ECG signal. The deep learning approaches such as 1D CNN and GRU can be helpful tools to automatically detect SA in sleep apnea screening and related studies.
Keywords: Sleep apnea | Deep learning | Convolutional neural network | Recurrent neural network | Long short-term memory | Gated-recurrent unit