Prediction of financial markets is a challenging task due to its volatility and the presence of noise. This work comparatively analyses the implementations of LSTM (Long Short Term Memory) and three kernels of SVR (Support Vector Regressor), namely linear kernel, RBF (Radial Basis Function) kernel and polynomial kernel. A series of experiments are conducted, and results and accuracy are compared. The results indicate that a single kernel of SVR is not sufficient to predict the stocks for all days, although if all the kernels are considered, one of them gets a close result but the LSTM with linear activation function and 4 layers give better result.

Performance comparison of clustering algorithms are often done in terms of different confusion matrix based scores obtained on test datasets when ground truth is available. However, a dataset comprises several instances having different difficulty levels. Therefore, it is more logical to compare effectiveness of clustering algorithms on individual instances instead of comparing scores obtained for the entire dataset. In this paper, an alternative approach is proposed for direct comparison of clustering algorithms in terms of individual instances within the dataset. A direct comparison matrix called Prayatul Matrix is prepared, which accounts for comparative outcome of two clustering algorithms on different instances of a dataset. Five different performance measures are designed based on prayatul matrix. Theoretical analysis shows proposed measures satisfy five important properties such as scale invariance, data invariance, permutation invariance,  monotonicity and continuity.  Efficacy of the proposed approach as well as  designed measures is analyzed empirically with four clustering algorithms on widely used standard datasets. Indications of proposed measures are compared with confusion matrix-based measures as well as other three permutation invariant measures. Results are evident that the newly designed measures are capable of giving some important insight about the clustering algorithms, which were impossible with the existing measures.

Study exploring the urban world deployment of LoRa waves using the ESP32  microcontroller Secure communication over large distances is not possible with conventional Wifi or cellular networks. A long range secure communication model is required for such conditions. This paper puts forth a communication model suitable for information exchange over a such large distances. LoRa (Long Range) is a spread spectrum modulation technique derived from chirp spread spectrum (CSS) technology. The typical range provided by LoRa can be up to three miles (5km) in urban areas, and up to 10 miles (15 km) or more in rural areas of line of sight. LoRa’s high range is characterized by high wireless link budgets of around 155 dB to 170 dB. ESP32 is a series of low-cost, low-power system on chip microcontrollers. The LoRa modem is available as the SX127X module which is connected to the ESP32 boards. Data security is ensured by performing data encryption on the sender board and data decryption on the reciever board. The algorithms for encryption and decryption shown in this article is the AES method. This paper also explores the possibility of hijacking the data while transmitting and shows which algorithms are immune to attacks and which can be easily broken. The use cases for such a communication system could be for secure 2-way communication, communication in remote areas where cellular and Wifi are not available, effecient control of IOT devices over a much larger distance, control of unmanned ground and aerial vehicles, etc.

This paper covers communication between planetary bodies in the solar system using a network of laser equipped-satellites. It proposes a novel communication protocol to minimize downtime and increase reliability in such a system in addition to a data synchronization system.

Trajectory Prediction (TP) is an important research topic in computer vision and robotics fields. Recently, many stochastic TP models have been proposed to deal with this problem and have achieved better performance than the traditional models with deterministic trajectory outputs. However, these stochastic models can generate a number of future trajectories with different qualities. They are lack of self-evaluation ability, that is, to examine the rationality of their prediction results, thus failing to guide users to identify high-quality ones from their candidate results. This hinders them from playing their best in real applications. In this paper, we make up for this defect and propose TPAD, a novel TP evaluation method based on the trajectory Anomaly Detection (AD) technique. In TPAD, we firstly combine the Automated Machine Learning (AutoML) technique and the experience in the AD and TP field to automatically design an effective trajectory AD model. Then, we utilize the learned trajectory AD model to examine the rationality of the predicted trajectories, and screen out good TP results for users. Extensive experimental results demonstrate that TPAD can effectively identify near-optimal prediction results, improving stochastic TP models’ practical application effect.

This review focuses on ultrasound beamforming methods, a crucial step in the process of ultrasound image formation. With the shift towards software-based receive beamforming in research and commercial settings, a number of limiting factors around the application of advanced beamforming algorithms have been removed.  Now, the opportunities for image quality improvements are only limited by available compute resources. This has lead to a plethora of works presenting and investigating novel beamforming methods. We review the recent advances in receive beamforming, specifically focused on methods that perform aperture data processing—the processing of the received echo signals after application of appropriate delays—and can therefore replace summation in delay-and-sum.  After presenting and contrasting the methods, we compare and discuss their respective advantages and limitations. Our goal is to give the reader an intuition on their working principles, show the relations between the different methods, and serve as a starting point for further study of the literature or implementations. The code we used to perform the comparisons is available at https://github.com/goeblr/rxbf_review

Smart contracts (SCs) are a set of logical procedures that can be run by individual peers participating within a Distributed Ledger Technology (DLT) network. By design, smart contracts inherit many of the benefits of DLT, including its immutability, scalability and  security properties. Nevertheless, they may introduce additional attack vectors, which can lead to cybersecurity explorations that could jeopardize the end-application’s ability to operate as intended or result in data leaks, and privacy violations. In this work an exploration of known problems, and possible attack scenarios will be presented. This is followed by a set of proposed best practices and mitigation strategies that are intended to assist developers, researchers and other relevant stakeholders to develop secure SC implementations.

In VANET, although data sharing among vehicles improves safety and traffic, malicious data providers might share false information to compromise its effectiveness. One typical solution to address this issue is to manage the trustworthiness of data providers by using feedback received from data users quantifying the veracity of the shared data. However, data users might behave maliciously and share untruthful feedback. This paper introduces TrCoin, a blockchain-based trust management system, for VANET which tackles both fake data and untruthful feedback. TrCoin uses an honesty value calculation algorithm to distinguish honest from malicious data users and filters out the feedback shared by the malicious data users. TrCoin then runs a trust value calculation algorithm to compute the trust values of data providers based on the truthful feedback shared by honest data users. Extensive simulations are conducted to demonstrate the effectiveness of TrCoin under various attack scenarios with false data injection and untruthful feedback.

In order to achieve improved tracking performance, this paper considers the distinct features of tailless flapping wing micro air vehicles (FWMAVs), and proposes a robust $\bm{H_{\infty}}$ controller. A linear parameter Varying (LPV) model is established for a tailless FWMAV with parameter uncertainty, based on the complete dynamic and kinetic model with quasi-steady and linear damping. Sampled data is used as feedback signals in controller design to overcome the influence of the control period not being much larger than the flapping period. An input delay technique is employed to convert the sampled-data system into a continuous-time system with a state delay. The controller is restricted to a designed multivariable proportional-integral (PI) structure for practical engineering implementation. A Lyapunov-Krasovskii function is employed to establish the stability and $\bm{H_{\infty}}$  performance of the closed-loop,  and an optimization problem involving linear matrix inequalities (LMIs) is obtained using diagonal Lyapunov matrix hypotheses. Numerical simulations demonstrate the strong robustness of the proposed controller, and flight experiments verify its effectiveness and superior performance compared to a state-of-the-art benchmark controller.

Session-based recommendation (SBR) systems aim to utilize the user’s short-term behavior sequence to predict the next item without the detailed user profile. Most recent works try to model the user preference by treating the sessions as between-item transition graphs and utilize various graph neural networks (GNNs) to encode the representations of pair-wise relations among items and their neighbors. Some of the existing GNN-based models mainly focus on aggregating information from the view of spatial graph structure, which ignores the temporal relations within neighbors of an item during message passing and the information loss results in a sub-optimal problem. Other works embrace this challenge by incorporating additional temporal information but lack sufficient interaction between the spatial and temporal patterns. To address this issue, inspired by the uniformity and alignment properties of contrastive learning techniques, we propose a novel framework called Session-based Recommendation with Spatio-Temporal Contrastive Learning Enhanced GNNs (RESTC). The idea is to supplement the GNN-based main supervised recommendation task with the temporal representation via an auxiliary cross-view contrastive learning mechanism. Furthermore, a novel global collaborative filtering graph (CFG) embedding is leveraged to enhance the spatial view in the main task. Extensive experiments demonstrate the significant performance of RESTC compared with the state-of-the-art baselines e.g., with an improvement as much as 27.08% gain on HR@20 and 20.10% gain on MRR@20.

Using computer simulations,  we  analyze the bit-error rate (BER) characteristics of a high-order QAM  optical transmission system repeated by two-mode phase-sensitive optical parametric amplifiers (PSAs).  In this  system, an idler, which is a phase-conjugated twin of a signal, propagate through the link together with the signal. One advantage of this system is that the  noise figure of two-mode PSAs used as repeaters can approach 0 dB. The other advantage is that  two-mode PSAs cancel out self-phase modulation of the signal at each repeater through the interaction between the transmitted  signal and idler. Owing to these advantages, the BER characteristics  of 16, 64, and 256QAM signals are significantly improved compared with the system using erbium-doped fiber amplifiers and phase-conjugation twin waves, provided that dispersion-shifted fibers are employed for transmission.

Using Newton’s gravitational equation, we have been able to calculate the orbital data of the planets in the solar system more accurately, but there are deviations between these data and astronomical observations. For example, the precession of Mercury’s orbit has a deviation of 43” per century. There are two important gravitational theories in modern physics: general relativity (GR) and quantum mechanics, of which GR has given an explanation for the precession deviation of Mercury, so from the perspective of quantum mechanics, can the quantum gravity equation also be derived? Can it be verified in the calculation of planetary orbits in the solar system? This paper will explore these questions.

The world is currently undergoing a rapid transformation in technology that will drastically change our lives, and potentially redefine what it means to be human. Machine learning has advanced significantly in the last few years. AI has become popular recently because of the large-scale data processing and managing capacities of machines. The application of machine learning in healthcare has two main domains: computer science and medical science. Machine learning techniques have brought advancement in medical science, allowing for the analysis of complex medical data. However, in the healthcare industry, machine learning serves as the doctor’s brain and knowledge. In this paper, we’ll look at some of the benefits of ML-based solutions and how they can be applied to healthcare. In recent years, many conferences have been held to explore new automated technologies in medicine to provide better health care. Several researchers are working in this domain to bring new innovations and features. This paper will outline the main advancements of machine learning in the automation of data analysis for patients’ health records and making predictions based on the data. It will also compare the leaps that have been made in computer-aided diagnosis, drug discovery, and personalized medicine.

An important component of an automated surveillance environment is person re-identification. The problem is often addressed using data received from vision sensors using appearance-based features, which are heavily reliant on visual cues such as colour, texture, and so on, limiting the reliability of re-identification of an individual. Much research has been performed to solve the problem of re-identification utilising human gait using inertial measurement units (IMU) data, which is thought to be unique and offer a distinct biometric signature that is especially ideal for re-ID in uncontrolled conditions. The locomotive activity of walking was the primary emphasis. The current study utilised not only locomotive activities but also non-locomotive activities of daily living. The data was obtained from the WISDM lab. The data is collected while engaging in six distinct everyday activities. The dataset was originally gathered for the purpose of Human Activity Recognition. Nonetheless, each person is given a unique ID. This information was utilised to re-identify the individual. The dataset consists of data of 36 volunteers. Shanakht-Net, a novel convolutional neural network, is introduced. The F1-score obtained is 93\%. Precision, recall, and accuracy are assessed and reported as well.

It is foreseeable that the 100 Gb/s/λ and beyond passive optical network (PON) will be required in future optical access networks to meet the explosive growth of data traffic. The coherent optical systems could be a promising solution for the future beyond 100G PON. Coherent PON using digital subcarrier multiplexing (DSCM) can provide flexible bandwidth allocation to a large number of access subscribers by dividing subcarriers of the DSCM signal into time slots for time-and-frequency division multiple access. When the optical network unit is allocated a new subcarrier, digital signal processing (DSP) should converge fast in the allocated time slot to ensure a low handoff latency for real-time bandwidth allocation. However, the traditional coherent DSP is hard to realize fast convergence due to blind and complex algorithms. In this paper, we design a specific training sequence (TS) structure and propose data-aided DSP to achieve fast convergence for coherent PON. The feasibility of the proposed scheme is experimentally verified in an  8 Gbaud/SC×8 SCs 400 Gb/s-net-rate coherent PON using DSCM with 16 quadrature amplitude modulation. The experimental results show that fast convergence is jointly realized by the proposed TS structure and data-aided DSP using a 416-symbol TS with a 52 ns duration. The receiver sensitivity at the 20% soft-decision forward error correction limit is approximately −27 dBm and an optical power budget of about 35.5 dB is achieved with a booster amplifier.

In this research paper, the problem of characterization of corner positive matrices ( an open research problem ) is addressed. Based on optimization of energy function associated with Hopfield Neural Network, an algorithmic approach to determine the corner positive definiteness of a symmetric matrix is proposed. It is reasoned that the problem addressed is an NP hard problem.

Grating-based X-ray phase-contrast and in particular dark-field radiography are promising new imaging modalities for medical applications. Currently, the potential advantage of dark-field imaging in early-stage diagnosis of pulmonary diseases in humans is being investigated. These studies make use of a comparatively large scanning interferometer at short acquisition times, which comes at the expense of a significantly reduced mechanical stability as compared to tabletop laboratory setups. Vibrations create random fluctuations of the grating alignment, causing artifacts in the resulting images. Here, we describe a novel maximum likelihood method for estimating this motion, thereby preventing these artifacts. It is tailored to scanning setups and does not require any sample-free areas. Unlike any previously described method, it accounts for motion in between as well as during exposures.

Paper on cryptography. The latest version of the paper has been accepted by Chinese Journal of Electronics.