Computer Science Journal (AGH University of Science and Technology, Krakow)
Not a member yet
476 research outputs found
Sort by
OPTIMIZED LOSSLESS AUDIO COMPRESSION USING DCT ENERGY THRESHOLDING AND MACHINE LEARNING TECHNIQUE
Full text linkIn this paper, a novel lossless audio compression technique has been proposed, utilizing the Discrete Cosine Transform (DCT) coefficient-controlled technique based on energy thresholding, an XOR-based neural network compression model, and a CNN model. Initially, the DCT is applied to the input audio signal to achieve better energy compaction, followed by transforming selected DCT coefficients into a compressed binary stream. Subsequently, this binary stream is passed to two prediction-based optimized models: an XOR model and a CNN model for further compression. The binary stream is first processed by the neural network model for XOR operation, and the resulting output is then fed into a CNN model to reduce data dimensionality and generate compressed audio data. The simulation findings are analyzed using various statistical and robustness measures and compared with existing approaches
DEVELOPING EXPLAINABLEMACHINE LEARNING MODEL USINGAUGMENTED CONCEPT ACTIVATION VECTOR
Full text linkMachine learning models use high-dimensional feature spaces to map their inputs to the corresponding class labels. However, these features often do not have a one-to-one correspondence with physical concepts understandable by humans, which hinders the ability to provide a meaningful explanation for the decisions made by these models. We propose a method for measuring the correlation between high-level concepts and the decisions made by a machine learning model. Our method can isolate the impact of a given high-level concept and accurately measure it quantitatively. Additionally, this study aims to determine the prevalence of frequent patterns in machine learning models, which often occur in imbalanced datasets. We have successfully applied the proposed method to fundus images and managed to quantitatively measure the impact of radiomic patterns on the model’s decisions
GDPKG-LLM: INTEGRATING GENE, DISEASE, AND PHARMACOGENOMICS KNOWLEDGE GRAPHS FOR COGNITIVE NEUROSCIENCE USING LARGE LANGUAGE MODELS
Full text linkUsing the structures of large language models (LLMs) in creating knowledgediagrams to understand more about the relationship between the entities ofcognitive and biological sciences has become a hot point of research. Due to thegreat knowledge behind the curtain and the deep connections of this research,it is not possible to use the traditional approaches of machine learning and deeplearning. In this study,the main goal is to create a comprehensive and integratedknowledge graph(KG) from the combination of three knowledge sources: GeneOntology (GO), Disease Ontology (DO), and PharmKG. Large language models(LLMs) have been used to create this knowledge base. The main purpose ofthis KG is to understand the relationships between genes, diseases and drugs.The pro- posed approach was called GDPKG-LLM. It has several key steps,including entity matching, similarity analysis, graph alignment and using GPT-4. GDPKG-LLM was able to extract more than 16,800 nodes and 838,000 edgesfrom these three knowledge bases and provide a rich KG. This graph providesmeaningful relationships, making it a valuable resource for future research inpersonalized medicine and neuroscience. The reviewed evaluation criteria showthe superiority of GDPKG-LLM, which strengthens the validity of this model
THE BENEFITS OF TESTING SOFTWARE IN SE RESEARCH: LESSONS LEARNED FROM TWO PhD PROJECTS
Full text linkSoftware engineering (SE) research often involves creating software, either as a primary research output (e.g. in design science research) or as a supporting tool to the traditional research process. Ensuring software quality is essential, as it influences both the research process and the credibility of findings. Integrating software testing methods into SE research can streamline efforts by addressing goals of both research and development processes simultaneously. This paper highlights the advantages of incorporating software testing in SE research, particularly for research evaluation. Through qualitative analysis of software artifacts and insights from two PhD projects, we present ten lessons learned. These experiences demonstrate that, when effectively integrated, software testing offers significant benefits for both the research process and its outcomes
TOWARD RAM FORENSICS SUPPORTEDBY MACHINE-LEARNING METHODS
Full text linkIn this article, we propose an enhancement to the computer forensics technique of using Machine Learning tools to analyse the contents of RAM in order to extract information potentially useful during an investigation. In the specific case presented, the use of the extracted information to generate more optimal dictionaries for dictionary cryptanalysis is considered. Increasing user awareness is making cryptanalysis of passwords increasingly difficult for law enforcement. Long and complex passwords are impossible to crack, even when high-performance computing platforms are available. A sensible method of optimization is to look for hints to use a dictionary that contains text phrases more likely to be used in the specific case under attack. Such a hint could be an analysis of RAM taken from the suspect computer. Machine learning methods can significantly facilitate this task. In this article, we also explore the effectiveness of such an approach and its usefulness in practical applications. We also consider applications of the proposed approach for other purposes, such as OSINT
DETECTION AND FORECASTING OF PARKINSON DISEASE PROGRESSION FROM SPEECH SIGNAL FEATURES USING MULTI-LAYER PERCEPTRON AND LSTM
Full text linkAccurate diagnosis of Parkinson disease, especially in its early stages, can be a challenging task. The application of machine learning techniques help improve the diagnostic accuracy of Parkinson’s disease progression. In this research work, two well-known feature selection methods (Relief-F and Sequential Forward Selection) were employed to identify the diagnostic features of audio signals of Parkinson disease patients and were used to train Multi-Layer Perceptron (MLP) and recurrent neural network Long Short-Term Memory(LSTM) for detection of disease and prediction of its progression. The MLP accurately detected Parkinson disease stages whereas LSTM successfully predicted Parkinson Stage 2 and 3
An improved context-aware Sentiment Analysis of student comments on Social Networks based on ChatGPT
Full text linkThe widespread use of social networks has provided a variety of active, dynamic, and popular platforms for students to express their opinions and sentiments. These data are increasingly being exploited and integrated into university information systems to better govern and manage universities and improve educational quality. The analysis of such data can offer valuable insights into student experiences and attitudes towards various educational aspects including courses, professors, events, and facilities. However, automatic opinion mining in this context is challenging due to the difficulty of analyzing some languages such as Arabic, the variety of used languages, the presence of informal language, the use of emoticons and emoji, sarcasm, and the need to consider the surrounding context. To deal with all these challenges, we propose a novel approach for an effective sentiment analysis of student comments on the X platform (Twitter). The proposed approach allows to collect student comments from Twitter public pages and automatically classifies comments into positive, negative, and neutral. The new approach is based on ChatGPT capabilities, supports three languages: English, Arabic, and colloquial Arabic, and integrates a new scoring method that measures both the positiveness and subjectivity of student comments. Experiments performed on simulated and real public Twitter pages of five Saudi high education institutions showed the performance of the proposed tool to automatically analyze and summarize collected data
A Physical Model of Quantum Bit Behavior Based on a Programmable FPGA Integrated Circuit
Full text linkThe rapidly developing field of quantum computing and the ongoing lack of widely available quantum computers create the need for scientists to build their simulators. However, mathematical simulation of such circuits usually ignores many aspects and problems found in real quantum systems. In this article, the authors describe a quantum bit emulator based on FPGA integrated circuits. In this case, FPGA technology provides real-time massive parallelism of the modeled physical phenomena. The modeled QUBIT is represented using a Bloch sphere. Its quantum state is set and modified only by precise pulses of an electrical signal, and with the help of similar pulses, it manifests its current state in real time. The constructed QUBIT was additionally equipped with decoherence mechanisms and with circuits that intentionally respond to internal and external noises that distort its current quantum state. This article presents and discusses how such a physically built emulator works
AQMLATOR – AN AUTO QUANTUM MACHINE LEARNING E-PLATFORM
Full text linkA successful Machine Learning (ML) model implementation requires threemain components: training dataset, suitable model architecture and trainingprocedure. Given dataset and task, finding an appropriate model might be challenging.AutoML, a branch of ML, focuses on automatic architecture search —a meta method that aims at removing human from ML system design process.The success of ML and the development of quantum computing (QC) in recentyears led to a birth of new fascinating field called Quantum Machine Learning(QML) that, amongst others, incorporates quantum computers into ML models.In this paper we present AQMLator, an Auto Quantum Machine Learningplatform that aims to automatically propose and train the quantum layers ofan ML model with minimal input from the user. This way, data scientists canbypass the entry barrier for QC and use QML. AQMLator uses standard MLlibraries, making it easy to introduce into existing ML pipelines
Quantum Inspired Chaotic Salp Swarm Optimization for Dynamic Optimization
Full text linkMany real-world problems are dynamic optimization problems that are unknown beforehand. In practice, unpredictable events such as the arrival of new jobs, due date changes, and reservation cancellations, changes in parameters or constraints make the search environment dynamic. Many algorithms are designed to deal with stationary optimization problems, but these algorithms do not face dynamic optimization problems or manage them correctly. Although some of the optimization algorithms are proposed to deal with the changes in dynamic environments differently, there are still areas of improvement in existing algorithms due to limitations or drawbacks, especially in terms of locating and following the previously identified optima. With this in mind, we studied a variant of SSA known as QSSO, which is integrating the principles of quantum computing. An attempt is made to improve the overall performance of standard SSA to deal with the dynamic environment effectively by locating and tracking the global optima for DOPs. This work is an extension of the proposed new algorithm QSSO, known as the Quantum-inspired Chaotic Salp Swarm Optimization (QCSSO) Algorithm, which is detailing the various approaches taken into consideration while solving DOPs. A chaotic operator is employed with quantum computing to respond to change and guarantee to increase individual searchability by improving population diversity and the speed at which the algorithm converges. We experimented by evaluating QCSSO on a well-known generalized dynamic benchmark problem (GDBG) provided for CEC 2009, followed by a comparative numerical study with well-regarded algorithms. As promised, the introduced QCSSO is discovered, and a rival algorithm for DOPs