INTERNATIONAL JOURNAL OF COMPUTERS & TECHNOLOGY

Positive-Unlabeled Learning Method for Positive Emotion Recognition Using EEG technology

Zizhu Li, Chengyuan Shen, Jianting Cao — Fri, 09 Aug 2024 00:00:00 +0000

Emotion is a reaction of the human brain to external events, and the study of emotion recognition has substantial practical applications. Therefore, accurately recognizing and understanding positive emotions across different populations is crucial. Traditional image recognition technology cannot effectively identify emotions in individuals with impaired facial muscle control, such as elderly people in nursing homes with Alzheimer’s disease and patients with facial nerve paralysis (Bell’s palsy). Consequently, many machine learning methods have been widely applied to emotion recognition based on electroencephalogram (EEG) signals in recent years. In cases where the number of samples is sufficient, powerful deep learning methods can achieve high performance in emotion recognition. However, obtaining a large amount of reliably labeled emotional EEG data is arduous. We introduce a Positive-Unlabeled (PU) learning method for classifying EEG signals into Positive and Non-Positive emotions using a binary classifier developed with minimal labeled data. This approach utilizes a small volume of labeled data containing only positive emotion signals, combined with unlabeled data that includes both classes, effectively reducing the dependency on extensive, reliably labeled EEG data. The best accuracy achieved by this method is 93.95%. Experimental results on the dataset demonstrate the
effectiveness of our approach.

A multidimensional EEG feature extraction attention detection for BCI System

Ran Zhang, Chengyuan Shen, Linfeng Sui, Professor Cao — Sat, 22 Jun 2024 00:00:00 +0000

Brain-computer interface (BCI) systems are highly valued for their applications in medicine and biology, where accurate analysis of attention levels and electroencephalography (EEG) data is crucial for the success rate of BCI systems. To improve this aspect, we propose to perform attention level determination prior to the execution of the BCI system. To accurately determine the level of attention, we propose a novel feature extraction approach that involves multidimensional feature extraction across multiple frequency bands of EEG data. The multi frequency band waveforms extracted using this method can be cross validated, thereby increasing the robustness of our research. We used publicly available datasets to train a Support Vector Machine (SVM) classifier to develop an efficient attention detection system, and developed a system for collecting EEG data to validate attention levels. The effectiveness and accuracy of the multidimensional feature extraction method was validated by classifying the data collected by the attention detection system. This study highlights the potential of integrating attention detection into BCI systems, pathways for advances in brain science research.

An EEG-based Sleep Staging method with hybrid entropy computation measures

Yiqian Yang, Shuchen Fu, Ruixiang Liao — Mon, 17 Jun 2024 00:00:00 +0000

Sleep is an indispensable physiological need of the human body. Sleep staging is an effective method to objectively assess sleep quality and is helpful for research on sleep and sleep-related diseases. Electroencephalogram (EEG) signals are nonlinear and non-stationary time series, and entropy features are particularly sensitive to these nonlinear characteristics and can reveal information that is difficult to discover with traditional linear analysis methods. We proposed an automatic sleep staging method based on EEG entropy computation, inlcuding signal preprocessing, entropy feature extraction, feature selection and lassification modules. The experimental results show that the average accruacy is 91.3% through the fused entropy features.

Redundant Residue Number System (RRNS) Di-Base Table for SOLiD Sequencing

Joshua Apigagua Akanbasiam, Kwame Osei Boateng, Matthew Glover Addo — Mon, 17 Jun 2024 00:00:00 +0000

The next-generation sequencing (NGS) methodology, sequencing by oligonucleotide ligation and detection (SOLiD) uses a di-base table, a somewhat unusual method, to decode sequences. Its coding scheme is based on the binary number system. The di-base table is not connected to the genetic code, nor is the coding scheme structured in the space of an entire number system. Gamow also revealed the hidden attribute of a 4 × 4 code for the di-base table, supporting his proposal of a 4 × 4 × 4 codons for the genetic code. Consideration for digital applications has focused more on the Residue Number System (RNS) and Redundant Residue Number System (RRNS) lately. Consequently, an RRNS di-base table based on the number tree concept is designed. The designed RRNS di-base table deviates from the canonical di-base table but retains every attribute necessary for effective SOLiD decoding. It shares a close relationship with the RNS-Genetic code and this presents a compelling argument for creating a single instrument that possesses the capabilities of both the genetic code and the di-base table.

Unveiling Neurophysiological Markers of Consciousness Levels through EEG Exploration

Wed, 05 Jun 2024 00:00:00 +0000

The concept of consciousness levels typically refers to various aspects and tiers related to an individual’s cognition, perception, thinking, and awareness. Although neurophysiological markers have not yet been definitively identified to distinguish between these nuanced levels, this paper introduces a robust marker, the Approximate Entropy (ApEn), which quantifies the complexity of EEG signals to differentiate states of altered consciousness. Utilizing ApEn, we analyze EEG data from the frontal lobe—a region closely associated with consciousness—in states indicative of severely altered conditions, specifically anesthesia, coma, and brain death. To enhance the precision of consciousness level
assessment, we employ a Support Vector Machine (SVM) model, which classifies the states based on EEG complexity measures. This approach not only provides valuable insights into the neural correlations associated with changes in these critical states but also underscores the potential of combining quantitative EEG analysis with machine learning techniques to advance our understanding of consciousness. The findings demonstrate that EEG complexity, when analyzed using ApEn coupled with SVM classification, offers a novel and effective method for assessing and distinguishing between degrees of consciousness. This approach promises significant implications for clinical diagnostics and patient monitoring.

A NEW ROBUST HOMOMORPHIC ENCRYPTION SCHEME BASED ON PAILLIER, RESIDUE NUMBER SYSTEM AND EL-GAMAL

Wed, 17 Apr 2024 00:00:00 +0000

The new focus of cryptographic research is on encryption schemes that can withstand cyber-attacks, with the arrival of cloud computing. The widely used public key encryption system designed by Taher El Gamal based on the discrete logarithm problem has been used in many sectors such as internet security, E-voting systems, and other applications for a long time. However, considering the potential data security threats in cloud computing, cryptologists are developing new and more robust cryptographic algorithms. To this end, a new robust homomorphic encryption scheme based on Paillier, Residue Number system (RNS), and El Gamal (PRE), is proposed in this paper., which is expected to be highly effective and resistant to cyber-attacks. The proposed scheme is composed a three-layer encryption and a three-layer decryption processes thereby, making it robust. It employs an existing RNS moduli set {2n + 1, 2n, 2n − 1, 2n−1} − 1}, having passed it through the Paillier encryption process for forward conversion and then the El Gamal cryptosystem to encrpyt any data. The decryption process is a reversal of these processes starting from the El Gamal through a reverse conversion with the same moduli set using the Chinese Remainder Theorem (CRT). The simulation results shows that the proposed scheme outperforms similar existing schemes in terms of robustness and therefore, making it more secured which however, trades off with the time of execution in similar comparison.

On Defining Smart Cities using Transformer Neural Networks

Andrei Khurshudov — Sun, 28 Jan 2024 00:00:00 +0000

Cities worldwide are rapidly adopting “smart” technologies, transforming urban life. Despite this trend, a universally accepted definition of “smart city” remains elusive. Past efforts to define it haven’t yielded a consensus, as evidenced by the numerous definitions in use. In this paper, we endeavored to create a new “compromise” definition that should resonate with most experts previously involved in defining this concept and aimed to validate one of the existing definitions. We reviewed 60 definitions of smart cities from industry, academia, and various relevant organizations, employing transformer architecture-based generative AI and semantic text analysis to reach this compromise. We proposed a semantic similarity measure as an evaluation technique, which could generally be used to compare different smart city definitions, assessing their uniqueness or resemblance. Our methodology employed generative AI to analyze various existing definitions of smart cities, generating a list of potential new composite definitions. Each of these new definitions was then tested against the pre-existing individual definitions we’ve gathered, using cosine similarity as our metric. This process identified smart city definitions with the highest average cosine similarity, semantically positioning them as the closest on average to all the 60 individual definitions selected.

Convolutional Neural Networks for Deep Sleep Detection Based on Data Augmentation

Ruixuan Chen, Linfeng Sui, Mo Xia, Jinsha Liu, Tao Zhang, Jianting Cao — Sun, 28 Jan 2024 00:00:00 +0000

Sleep is a necessary process that individuals undergo daily for physical recovery, and the proportion of deep sleep in the sleep stages is a critical aspect of the recovery process. Convolutional Neural Networks (CNNs) have shown remarkable success in automatically identifying deep sleep stages through the analysis of electroencephalogram (EEG) signals. This article introduces three data augmentation techniques, including time shifting, amplitude scaling and noise addition, to enhance the diversity and features of the data. These techniques aim to enable machine learning models to extract features from various aspects of sleep EEG data, thus improving the model’s accuracy. Three deep learning models are introduced, namely DeepConvNet, ShallowConvNet and EEGNet, for the identification of deep sleep. To evaluate the proposed methods, the Sleep-EDF public dataset was utilized. Experimental results demonstrate that the enhanced dataset formed by applying the three data augmentation techniques achieved higher accuracy in all deep learning models compared to the original dataset. This highlights the feasibility and effectiveness of these methods in deep sleep detection.