ALBERT

Beschreibung: Learned image reconstruction techniques using deep neural networks have recently gained popularity and have delivered promising empirical results. However, most approaches focus on one single recovery for each observation, and thus neglect information uncertainty. In this work, we develop a novel computational framework that approximates the posterior distribution of the unknown image at each query observation. The proposed framework is very flexible: it handles implicit noise models and priors, it incorporates the data formation process (i.e., the forward operator), and the learned reconstructive properties are transferable between different datasets. Once the network is trained using the conditional variational autoencoder loss, it provides a computationally efficient sampler for the approximate posterior distribution via feed-forward propagation, and the summarizing statistics of the generated samples are used for both point-estimation and uncertainty quantification. We illustrate the proposed framework with extensive numerical experiments on positron emission tomography (with both moderate and low-count levels) showing that the framework generates high-quality samples when compared with state-of-the-art methods.

Beschreibung: Brain tumor segmentation seeks to separate healthy tissue from tumorous regions. This is an essential step in diagnosis and treatment planning to maximize the likelihood of successful treatment. Magnetic resonance imaging (MRI) provides detailed information about brain tumor anatomy, making it an important tool for effective diagnosis which is requisite to replace the existing manual detection system where patients rely on the skills and expertise of a human. In order to solve this problem, a brain tumor segmentation & detection system is proposed where experiments are tested on the collected BraTS 2018 dataset. This dataset contains four different MRI modalities for each patient as T1, T2, T1Gd, and FLAIR, and as an outcome, a segmented image and ground truth of tumor segmentation, i.e., class label, is provided. A fully automatic methodology to handle the task of segmentation of gliomas in pre-operative MRI scans is developed using a U-Net-based deep learning model. The first step is to transform input image data, which is further processed through various techniques—subset division, narrow object region, category brain slicing, watershed algorithm, and feature scaling was done. All these steps are implied before entering data into the U-Net Deep learning model. The U-Net Deep learning model is used to perform pixel label segmentation on the segment tumor region. The algorithm reached high-performance accuracy on the BraTS 2018 training, validation, as well as testing dataset. The proposed model achieved a dice coefficient of 0.9815, 0.9844, 0.9804, and 0.9954 on the testing dataset for sets HGG-1, HGG-2, HGG-3, and LGG-1, respectively.

Beschreibung: The two-part series of papers provides a survey on recent advances in Deep Reinforcement Learning (DRL) for solving partially observable Markov decision processes (POMDP) problems. Reinforcement Learning (RL) is an approach to simulate the human’s natural learning process, whose key is to let the agent learn by interacting with the stochastic environment. The fact that the agent has limited access to the information of the environment enables AI to be applied efficiently in most fields that require self-learning. It’s essential to have an organized investigation—we can make good comparisons and choose the best structures or algorithms when applying DRL in various applications. The first part of the overview introduces Markov Decision Processes (MDP) problems and Reinforcement Learning and applications of DRL for solving POMDP problems in games, robotics, and natural language processing. In part two, we continue to introduce applications in transportation, industries, communications and networking, etc. and discuss the limitations of DRL.

Beschreibung: The learning with errors (LWE) problem is one of the main mathematical foundations of post-quantum cryptography. One of the main groups of algorithms for solving LWE is the Blum–Kalai–Wasserman (BKW) algorithm. This paper presents new improvements of BKW-style algorithms for solving LWE instances. We target minimum concrete complexity, and we introduce a new reduction step where we partially reduce the last position in an iteration and finish the reduction in the next iteration, allowing non-integer step sizes. We also introduce a new procedure in the secret recovery by mapping the problem to binary problems and applying the fast Walsh Hadamard transform. The complexity of the resulting algorithm compares favorably with all other previous approaches, including lattice sieving. We additionally show the steps of implementing the approach for large LWE problem instances. We provide two implementations of the algorithm, one RAM-based approach that is optimized for speed, and one file-based approach which overcomes RAM limitations by using file-based storage.

Beschreibung: In this paper, a discrete moth–flame optimization algorithm for community detection (DMFO-CD) is proposed. The representation of solution vectors, initialization, and movement strategy of the continuous moth–flame optimization are purposely adapted in DMFO-CD such that it can solve the discrete community detection. In this adaptation, locus-based adjacency representation is used to represent the position of moths and flames, and the initialization process is performed by considering the community structure and the relation between nodes without the need of any knowledge about the number of communities. Solution vectors are updated by the adapted movement strategy using a single-point crossover to distance imitating, a two-point crossover to calculate the movement, and a single-point neighbor-based mutation that can enhance the exploration and balance exploration and exploitation. The fitness function is also defined based on modularity. The performance of DMFO-CD was evaluated on eleven real-world networks, and the obtained results were compared with five well-known algorithms in community detection, including GA-Net, DPSO-PDM, GACD, EGACD, and DECS in terms of modularity, NMI, and the number of detected communities. Additionally, the obtained results were statistically analyzed by the Wilcoxon signed-rank and Friedman tests. In the comparison with other comparative algorithms, the results show that the proposed DMFO-CD is competitive to detect the correct number of communities with high modularity.

Beschreibung: Solving Constraint Optimization Problems (COPs) can be dramatically simplified by boundary estimation, that is providing tight boundaries of cost functions. By feeding a supervised Machine Learning (ML) model with data composed of the known boundaries and extracted features of COPs, it is possible to train the model to estimate the boundaries of a new COP instance. In this paper, we first give an overview of the existing body of knowledge on ML for Constraint Programming (CP), which learns from problem instances. Second, we introduce a boundary estimation framework that is applied as a tool to support a CP solver. Within this framework, different ML models are discussed and evaluated regarding their suitability for boundary estimation, and countermeasures to avoid unfeasible estimations that avoid the solver finding an optimal solution are shown. Third, we present an experimental study with distinct CP solvers on seven COPs. Our results show that near-optimal boundaries can be learned for these COPs with only little overhead. These estimated boundaries reduce the objective domain size by 60-88% and can help the solver find near-optimal solutions early during the search.

Beschreibung: Although visible light communication (VLC) channels are more secure than radio frequency channels, the broadcast nature of VLC links renders them open to eavesdropping. As a result, VLC networks must provide security in order to safeguard the user’s data from eavesdroppers. In the literature, keyless security techniques have been developed to offer security for VLC. Even though these techniques provide strong security against eavesdroppers, they are difficult to deploy. Key generation algorithms are critical for securing wireless connections. Nonetheless, in many situations, the typical key generation methods may be quite complicated and costly. They consume scarce resources, such as bandwidth. In this paper, we propose a novel key extraction procedure that uses error-correcting coding and one time pad (OTP) to improve the security of VLC networks and the validity of data. This system will not have any interference problems with other devices. We also explain error correction while sending a message across a network, and suggest a change to the Berlekamp–Massey (BM) algorithm for error identification and assessment. Because each OOK signal frame is encrypted by a different key, the proposed protocol provides high physical layer security; it allows for key extraction based on the messages sent, so an intruder can never break the encryption system, even if the latter knows the protocol with which we encrypted the message; our protocol also enables for error transmission rate correction and bit mismatch rates with on-the-fly key fetch. The results presented in this paper were performed using MATLAB.

Beschreibung: Text classification is an important component in many applications. Text classification has attracted the attention of researchers to continue to develop innovations and build new classification models that are sourced from clinical trial texts. In building classification models, many methods are used, including supervised learning. The purpose of this study is to improve the computational performance of one of the supervised learning methods, namely KNN, in building a clinical trial document text classification model by combining KNN and the fine-grained algorithm. This research contributed to increasing the computational performance of KNN from 388,274 s to 260,641 s in clinical trial texts on a clinical trial text dataset with a total of 1,000,000 data.

Beschreibung: An information retrieval (IR) system is the core of many applications, including digital library management systems (DLMS). The IR-based DLMS depends on either the title with keywords or content as symbolic strings. In contrast, it ignores the meaning of the content or what it indicates. Many researchers tried to improve IR systems either using the named entity recognition (NER) technique or the words’ meaning (word sense) and implemented the improvements with a specific language. However, they did not test the IR system using NER and word sense disambiguation together to study the behavior of this system in the presence of these techniques. This paper aims to improve the information retrieval system used by the DLMS by adding the NER and word sense disambiguation (WSD) together for the English and Arabic languages. For NER, a voting technique was used among three completely different classifiers: rules-based, conditional random field (CRF), and bidirectional LSTM-CNN. For WSD, an examples-based method was used to implement it for the first time with the English language. For the IR system, a vector space model (VSM) was used to test the information retrieval system, and it was tested on samples from the library of the University of Kufa for the Arabic and English languages. The overall system results show that the precision, recall, and F-measures were increased from 70.9%, 74.2%, and 72.5% to 89.7%, 91.5%, and 90.6% for the English language and from 66.3%, 69.7%, and 68.0% to 89.3%, 87.1%, and 88.2% for the Arabic language.

Beschreibung: In the last few years, a number of social media e-business models including the social networking giants of Facebook, Pinterest and Instagram have offered direct purchase abilities to both their users and the involved enterprises. Hence, individuals can buy directly without having to leave the social media website. At the same time, there is a significant increase in the number of online purchases through mobile devices. To add to this, nowadays, the vast majority of internet users prefer to surf via their smartphone rather than a desktop PC. The aforementioned facts reveal the abilities and potential dynamics of Mobile Social Commerce (MSC), which is considered not only the present but also the future of e-commerce, as well as an area of prosperous academic and managerial concern. In spite of its several extant abilities and its booming future, MSC has been little examined until now. Therefore, this study aims to determine the factors that impact smartphone users’ behavioral intention to adopt direct purchases through social media apps in a country where these kinds of m-services are not yet available. In specific, it extends the well-established Unified Theory of Acceptance and Use of Technology (UTAUT) model with the main ICT facilitators (i.e., convenience, reward and security) and inhibitors (i.e., risk and anxiety). The suggested conceptual model aims to increase the understanding on the topic and strengthen the importance of this major type of MSC. Convenience sampling was applied to gather the data and Structural Equation Modeling (SEM) was then performed to investigate the research hypotheses of the proposed conceptual model. The results show that performance expectancy exerts a positive impact on behavioral intention. Furthermore, all ICT facilitators examined do impact significantly on smartphone users’ decision to adopt direct mobile purchases through social media apps, whereas anxiety exerts a negative effect.