Search results for: fault detection and classification

Authors: S. B. Kumbalavati, J. D. Mallapur, K. Y. Bendigeri

Abstract:

A mobile Ad-hoc network (MANET) is a multihop wireless network where nodes communicate each other without any pre-deployed infrastructure. There is no central administrating unit. Hence, MANET is generally prone to many of the attacks. These attacks may alter, release or deny data. These attacks are nothing but intrusions. Intrusion is a set of actions that attempts to compromise integrity, confidentiality and availability of resources. A major issue in the design and operation of ad-hoc network is sharing the common spectrum or common channel bandwidth among all the nodes. We are performing intrusion detection using game theory approach. Game theory is a mathematical tool for analysing problems of competition and negotiation among the players in any field like marketing, e-commerce and networking. In this paper mathematical model is developed using game theory approach and intruders are detected and removed. Bandwidth utilization is estimated and comparison is made between bandwidth utilization with intrusion detection technique and without intrusion detection technique. Percentage of intruders and efficiency of the network is analysed.

Keywords: ad-hoc network, IDS, game theory, sensor networks

Procedia PDF Downloads 387

Authors: Sehreen Moorat, Hiba, Maham Mahnoor, Faryal Soomro

Abstract:

Leakage of gases in a system makes infrastructures and users vulnerable; it can occur due to its environmental conditions or old groundwork. In hospitals and industries, it is very important to detect any small level of gas leakage because of their sensitivity. In this research, a portable detection system for the small leakage of gases has been developed, gas sensor (MQ-2) is used to find leakage when it’s at its initial phase. The sensor and transmitting module senses the change in level of gas by using a sensing circuit. When a concentration of gas reach at a specified threshold level, it will activate an alarm and send the alarming situation notification to receiver through GSM module. The proposed system works well in hospitals, home, and industries.

Keywords: gases, detection, Arduino, MQ-2, alarm

Procedia PDF Downloads 206

Authors: Cyrus Addy, Venkata Sriram Siddhardh Nadendla, Kwame Awuah-Offei

Abstract:

Surface robots in modern underground mine rescue operations suffer from several limitations in enabling a prompt self-rescue. Therefore, the possibility of designing and deploying in-mine robots to expedite miner self-rescue can have a transformative impact on miner safety. These in-mine robots for miner self-rescue can be envisioned to carry out diverse tasks such as object detection, autonomous navigation, and payload delivery. Specifically, this paper investigates the challenges in the design of object detection algorithms for in-mine robots using thermal images, especially to detect people in real-time. A total of 125 thermal images were collected in the Missouri S&T Experimental Mine with the help of student volunteers using the FLIR TG 297 infrared camera, which were pre-processed into training and validation datasets with 100 and 25 images, respectively. Three state-of-the-art, pre-trained real-time object detection models, namely YOLOv5, YOLO-FIRI, and YOLOv8, were considered and re-trained using transfer learning techniques on the training dataset. On the validation dataset, the re-trained YOLOv8 outperforms the re-trained versions of both YOLOv5, and YOLO-FIRI.

Keywords: miner self-rescue, object detection, underground mine, YOLO

Procedia PDF Downloads 83

Authors: Anne-Lena Kampen, Øivind Kure

Abstract:

Machine learning (ML) can be implemented in Wireless Sensor Networks (WSNs) as a central solution or distributed solution where the ML is embedded in the nodes. Embedding improves privacy and may reduce prediction delay. In addition, the number of transmissions is reduced. However, quality factors such as prediction accuracy, fault detection efficiency and coordinated control of the overall system suffer. Here, we discuss and highlight the trade-offs that should be considered when choosing between embedding and centralized ML, especially for multihop networks. In addition, we present estimations that demonstrate the energy trade-offs between embedded and centralized ML. Although the total network energy consumption is lower with central prediction, it makes the network more prone for partitioning due to the high forwarding load on the one-hop nodes. Moreover, the continuous improvements in the number of operations per joule for embedded devices will move the energy balance toward embedded prediction.

Keywords: central machine learning, embedded machine learning, energy consumption, local machine learning, wireless sensor networks, WSN

Procedia PDF Downloads 153

Authors: Krisala Joshi, Sayanta Ghosh, Renu Lata, Jagdish C. Kuniyal

Abstract:

Throughout the world, monitoring and estimating the changing pattern of forests across diverse landscapes through remote sensing is instrumental in understanding the interactions of human activities and the ecological environment with the changing climate. Forest change detection using satellite imageries has emerged as an important means to gather information on a regional scale. Kullu valley in Himachal Pradesh, India is situated in a transitional zone between the lesser and the greater Himalayas. Thus, it presents a typical rugged mountainous terrain with moderate to high altitude which varies from 1200 meters to over 6000 meters. Due to changes in agricultural cropping patterns, urbanization, industrialization, hydropower generation, climate change, tourism, and anthropogenic forest fire, it has undergone a tremendous transformation in forest cover in the past three decades. The loss and degradation of forest cover results in soil erosion, loss of biodiversity including damage to wildlife habitats, and degradation of watershed areas, and deterioration of the overall quality of nature and life. The supervised classification of LANDSAT satellite data was performed to assess the changes in forest cover in Kullu valley over the years 2000 to 2020. Normalized Burn Ratio (NBR) was calculated to discriminate between burned and unburned areas of the forest. Our study reveals that in Kullu valley, the increasing number of forest fire incidents specifically, those due to anthropogenic activities has been on a rise, each subsequent year. The main objective of the present study is, therefore, to estimate the change in the forest cover of Kullu valley and to address the various social aspects responsible for the anthropogenic forest fires. Also, to assess its impact on the significant changes in the regional climatic factors, specifically, temperature, humidity, and precipitation over three decades, with the help of satellite imageries and ground data. The main outcome of the paper, we believe, will be helpful for the administration for making a quantitative assessment of the forest cover area changes due to anthropogenic activities and devising long-term measures for creating awareness among the local people of the area.

Keywords: Anthropogenic Activities, Forest Change Detection, Normalized Burn Ratio (NBR), Supervised Classification

Procedia PDF Downloads 173

Authors: Sondos M. Hammad, Sherin M. ElGokhy, Mahmoud M. Fahmy, Elsayed A. Sallam

Abstract:

In this paper, an ensemble-based Support Vector Machine (SVM) classification approach is proposed. It is used for miRNA prediction. Three problems, commonly associated with previous approaches, are alleviated. These problems arise due to impose assumptions on the secondary structural of premiRNA, imbalance between the numbers of the laboratory checked miRNAs and the pseudo-hairpins, and finally using a training data set that does not consider all the varieties of samples in different species. We aggregate the predicted outputs of three well-known SVM classifiers; namely, Triplet-SVM, Virgo and Mirident, weighted by their variant features without any structural assumptions. An additional SVM layer is used in aggregating the final output. The proposed approach is trained and then tested with balanced data sets. The results of the proposed approach outperform the three base classifiers. Improved values for the metrics of 88.88% f-score, 92.73% accuracy, 90.64% precision, 96.64% specificity, 87.2% sensitivity, and the area under the ROC curve is 0.91 are achieved.

Keywords: MiRNAs, SVM classification, ensemble algorithm, assumption problem, imbalance data

Procedia PDF Downloads 349

Authors: Sulaiman Al Amro

Abstract:

There are currently considerable challenges concerning data security and privacy, particularly in relation to modern technologies. This includes the virtual world known as the Metaverse, which consists of a virtual space that integrates various technologies and is therefore susceptible to cyber threats such as malware, phishing, and identity theft. This has led recent studies to propose the development of Metaverse forensic frameworks and the integration of advanced technologies, including machine learning for intrusion detection and security. In this context, the application of first-order logic offers a formal and systematic approach to defining the conditions of cyberattacks, thereby contributing to the development of effective detection mechanisms. In addition, formalizing the rules and patterns of cyber threats has the potential to enhance the overall security posture of the Metaverse and, thus, the integrity and safety of this virtual environment. The current paper focuses on the primary actions employed by avatars for potential attacks, including Interval Temporal Logic (ITL) and behavior-based detection to detect an avatar’s abnormal activities within the Metaverse. The research established that the proposed framework attained an accuracy of 92.307%, resulting in the experimental results demonstrating the efficacy of ITL, including its superior performance in addressing the threats posed by avatars within the Metaverse domain.

Keywords: security, privacy, metaverse, cyberattacks, detection, first-order logic

Procedia PDF Downloads 41

Authors: Babak Kamrani, Nozar Kishi

Abstract:

Stochastic catalog represents the events module of the earthquake loss estimation models. It includes series of events with different magnitudes and corresponding frequencies/probabilities. For the development of the stochastic catalog, random or uniform sampling methods are used to sample the events from the seismicity model. For covering all the Magnitude Frequency Distribution (MFD), a huge number of events should be generated for the above-mentioned methods. Characteristic Event (CE) method chooses the events based on the interest of the insurance industry. We divide the MFD of each source into bins. We have chosen the bins based on the probability of the interest by the insurance industry. First, we have collected the information for the available seismic sources. Sources are divided into Fault sources, subduction, and events without specific fault source. We have developed the MFD for each of the individual and areal source based on the seismicity of the sources. Afterward, we have calculated the CE magnitudes based on the desired probability. To develop the stochastic catalog, we have introduced uncertainty to the location of the events too.

Keywords: stochastic catalogue, earthquake loss, uncertainty, characteristic event

Procedia PDF Downloads 300

Authors: Deepanjali Gurav, Kun Qian

Abstract:

In-vitro metabolic diagnosis relies on designed materials-based analytical platforms for detection of selected metabolites in biological samples, which has a key role in disease detection and therapeutic evaluation in clinics. However, the basic challenge deals with developing a simple approach for metabolic analysis in bio-samples with high sample complexity and low molecular abundance. In this work, we report a designer plasmonic nanoshells based platform for direct detection of small metabolites in clinical samples for in-vitro metabolic diagnostics. We first synthesized a series of plasmonic core-shell particles with tunable nanoshell structures. The optimized plasmonic nanoshells as new matrices allowed fast, multiplex, sensitive, and selective LDI MS (Laser desorption/ionization mass spectrometry) detection of small metabolites in 0.5 μL of bio-fluids without enrichment or purification. Furthermore, coupling with isotopic quantification of selected metabolites, we demonstrated the use of these plasmonic nanoshells for disease detection and therapeutic evaluation in clinics. For disease detection, we identified patients with postoperative brain infection through glucose quantitation and daily monitoring by cerebrospinal fluid (CSF) analysis. For therapeutic evaluation, we investigated drug distribution in blood and CSF systems and validated the function and permeability of blood-brain/CSF-barriers, during therapeutic treatment of patients with cerebral edema for pharmacokinetic study. Our work sheds light on the design of materials for high-performance metabolic analysis and precision diagnostics in real cases.

Keywords: plasmonic nanoparticles, metabolites, fingerprinting, mass spectrometry, in-vitro diagnostics

Procedia PDF Downloads 138

Authors: Cuneyt Yucelbas, Seral Ozsen, Sule Yucelbas, Gulay Tezel

Abstract:

Due to the fact that there exist only a small number of complex systems in artificial immune system (AIS) that work out nonlinear problems, nonlinear AIS approaches, among the well-known solution techniques, need to be developed. Gaussian function is usually used as similarity estimation in classification problems and pattern recognition. In this study, diagnosis of breast cancer, the second type of the most widespread cancer in women, was performed with different distance calculation functions that euclidean, gaussian and gaussian-euclidean hybrid function in the clonal selection model of classical AIS on Wisconsin Breast Cancer Dataset (WBCD), which was taken from the University of California, Irvine Machine-Learning Repository. We used 3-fold cross validation method to train and test the dataset. According to the results, the maximum test classification accuracy was reported as 97.35% by using of gaussian-euclidean hybrid function for fold-3. Also, mean of test classification accuracies for all of functions were obtained as 94.78%, 94.45% and 95.31% with use of euclidean, gaussian and gaussian-euclidean, respectively. With these results, gaussian-euclidean hybrid function seems to be a potential distance calculation method, and it may be considered as an alternative distance calculation method for hard nonlinear classification problems.

Keywords: artificial immune system, breast cancer diagnosis, Euclidean function, Gaussian function

Procedia PDF Downloads 435

Authors: Muhammad Ameer Nawaz Akram

Abstract:

The 21st century has been revealed that many individuals around the world are living in urban settlements than in rural zones. The evolution of numerous cities in emerging and newly developed countries is accompanied by the rise of slums. The precise definition of a slum varies countries to countries, but the universal harmony is that slums are dilapidated settlements facing severe poverty and have lacked access to sanitation, water, electricity, good living styles, and land tenure. The slum settlements always vary in unique patterns within and among the countries and cities. The core objective of this study is the spatial identification and classification of slums in Gujrat city Pakistan from very high-resolution GeoEye-1 (0.41m) satellite imagery. Slums were first identified using GPS for sample site identification and ground-truthing; through this process, 425 slums were identified. Then Object-Oriented Analysis (OOA) was applied to classify slums on digital image. Spatial analysis softwares, e.g., ArcGIS 10.3, Erdas Imagine 9.3, and Envi 5.1, were used for processing data and performing the analysis. Results show that OOA provides up to 90% accuracy for the identification of slums. Jalal Cheema and Allah Ho colonies are severely affected by slum settlements. The ratio of criminal activities is also higher here than in other areas. Slums are increasing with the passage of time in urban areas, and they will be like a hazardous problem in coming future. So now, the executive bodies need to make effective policies and move towards the amelioration process of the city.

Keywords: slums, GPS, satellite imagery, object oriented analysis, zonal change detection

Procedia PDF Downloads 134

Authors: Neha Ahirwar

Abstract:

In the contemporary digital era, the rise of credit card fraud poses a significant threat to both financial institutions and consumers. As fraudulent activities become more sophisticated, there is an escalating demand for robust and effective fraud detection mechanisms. Advanced machine learning algorithms have become crucial tools in addressing this challenge. This paper conducts a thorough examination of the design and evaluation of a credit card fraud detection system, utilizing four prominent machine learning algorithms: random forest, logistic regression, decision tree, and XGBoost. The surge in digital transactions has opened avenues for fraudsters to exploit vulnerabilities within payment systems. Consequently, there is an urgent need for proactive and adaptable fraud detection systems. This study addresses this imperative by exploring the efficacy of machine learning algorithms in identifying fraudulent credit card transactions. The selection of random forest, logistic regression, decision tree, and XGBoost for scrutiny in this study is based on their documented effectiveness in diverse domains, particularly in credit card fraud detection. These algorithms are renowned for their capability to model intricate patterns and provide accurate predictions. Each algorithm is implemented and evaluated for its performance in a controlled environment, utilizing a diverse dataset comprising both genuine and fraudulent credit card transactions.

Keywords: efficient credit card fraud detection, random forest, logistic regression, XGBoost, decision tree

Procedia PDF Downloads 67

Authors: Waroton Paisuwan, Mongkol Sukwattanasinitt, Mamoru Tobisu, Anawat Ajavakom

Abstract:

Novel dihydroquinoline derivatives (DHP and DHP-OH) were synthesized in one pot via a tandem trimerization-cyclization of methylpropiolate. DHP and DHP-OH possess strong blue fluorescence with high quantum efficiencies over 0.70 in aqueous media. DHP-OH displays a remarkable fluorescence quenching selectively to the presence of Au3+ through the oxidation of dihydropyridine to pyridinium ion as confirmed by NMR and HRMS. DHP-OH was used to demonstrate the quantitative analysis of Au3+ in water samples with the limit of detection of 33 ppb and excellent recovery (>95%). This fluorescent probe was also applied for the determination of Au3+ residue in the gold nanoparticle solution and a paper-based sensing strip for the on-site detection of Au3+.

Keywords: Gold(III) ion detection, Fluorescent sensor, Fluorescence quenching, Dihydropyridine, Gold nanoparticles (AuNPs)

Procedia PDF Downloads 86

Authors: M. Malekian, M. E. Heydari, M. Irani Estyar

Abstract:

Human papillomavirus (HPV) is one of the most common sexually transmitted infection, which may lead to cervical cancer as the main cause of it. With early diagnosis and treatment in health care services, cervical cancer and its complications are considered to be preventable. This study was aimed to compare the efficiency, sensitivity, and specificity of Pap smear and polymerase chain reaction (PCR) in detecting HPV. A literature search was performed in Google Scholar, PubMed and SID databases using the keywords 'human papillomavirus', 'pap smear' and 'polymerase change reaction' to identify studies comparing Pap smear and PCR methods for the detection. No restrictions were considered.10 studies were included in this review. All samples that were positive by pop smear were also positive by PCR. However, there were positive samples detected by PCR which was negative by pop smear and in all studies, many positive samples were missed by pop smear technique. Although The Pap smear had high specificity, PCR based HPV detection was more sensitive method and had the highest sensitivity. In order to promote the quality of detection and high achievement of the maximum results, PCR diagnostic methods in addition to the Pap smear are needed and Pap smear method should be combined with PCR techniques according to the high error rate of Pap smear in detection.

Keywords: human papillomavirus, cervical cancer, pap smear, polymerase chain reaction

Procedia PDF Downloads 131

Authors: Rosa L. Figueroa, Christopher A. Flores, Qing Zeng-Treitler

Abstract:

A regular expression consists of sequence characters which allow describing a text path. Usually, in clinical research, regular expressions are manually created by programmers together with domain experts. Lately, there have been several efforts to investigate how to generate them automatically. This article presents a text classification algorithm based on regexes. The algorithm named REX was designed, and then, implemented as a simplified method to create regexes to classify Spanish text automatically. In order to classify ambiguous cases, such as, when multiple labels are assigned to a testing example, REX includes an information gain method Two sets of data were used to evaluate the algorithm’s effectiveness in clinical text classification tasks. The results indicate that the regular expression based classifier proposed in this work performs statically better regarding accuracy and F-measure than Support Vector Machine and Naïve Bayes for both datasets.

Keywords: information gain, regular expressions, smith-waterman algorithm, text classification

Procedia PDF Downloads 320

Authors: Md Fazle Alam, Amaj Ahmed Laskar, Hina Younus

Abstract:

Melamine toxicity which causes renal failure and death of humans and animals have recently attracted worldwide attention. Developing an easy, fast and sensitive method for the routine melamine detection is the need of the hour. Herein, we have developed a rapid, sensitive, one step and selective colorimetric method for the detection of melamine in milk samples based upon in-situ formation of silver nanoparticles (AgNPs) via tannic acid at room temperature. These AgNPs thus formed were characterized by UV-VIS spectrophotometer, transmission electron microscope (TEM), zetasizer and dynamic light scattering (DLS). Under optimal conditions, melamine could be selectively detected within the concentration range of 0.05-1.4 µM with a limit of detection (LOD) of 10.1 nM, which is lower than the strictest melamine safety requirement of 1 ppm. This assay does not utilize organic cosolvents, enzymatic reactions, light sensitive dye molecules and sophisticated instrumentation, thereby overcoming some of the limitations of conventional methods.

Keywords: milk adulteration, melamine, silver nanoparticles, tannic acid

Procedia PDF Downloads 246

Authors: Abe Degale D., Cheng Jian

Abstract:

When creating automated video surveillance systems, violent action recognition is crucial. In recent years, hand-crafted feature detectors have been the primary method for achieving violence detection, such as the recognition of fighting activity. Researchers have also looked into learning-based representational models. On benchmark datasets created especially for the detection of violent sequences in sports and movies, these methods produced good accuracy results. The Hockey dataset's videos with surveillance camera motion present challenges for these algorithms for learning discriminating features. Image recognition and human activity detection challenges have shown success with deep representation-based methods. For the purpose of detecting violent images and identifying aggressive human behaviours, this research suggested a deep representation-based model using the transfer learning idea. The results show that the suggested approach outperforms state-of-the-art accuracy levels by learning the most discriminating features, attaining 99.34% and 99.98% accuracy levels on the Hockey and Movies datasets, respectively.

Keywords: violence detection, faster RCNN, transfer learning and, surveillance video

Procedia PDF Downloads 108

Authors: Tolulope Aremu

Abstract:

The key process steps to produce liquid detergent products will introduce potential defects, such as formulation, mixing, filling, and packaging, which might compromise product quality, consumer safety, and operational efficiency. Real-time identification and characterization of such defects are of prime importance for maintaining high standards and reducing waste and costs. Usually, defect detection is performed by human inspection or rule-based systems, which is very time-consuming, inconsistent, and error-prone. The present study overcomes these limitations in dealing with optimization in defect characterization within the process for making liquid detergents using Machine Learning algorithms. Performance testing of various machine learning models was carried out: Support Vector Machine, Decision Trees, Random Forest, and Convolutional Neural Network on defect detection and classification of those defects like wrong viscosity, color deviations, improper filling of a bottle, packaging anomalies. These algorithms have significantly benefited from a variety of optimization techniques, including hyperparameter tuning and ensemble learning, in order to greatly improve detection accuracy while minimizing false positives. Equipped with a rich dataset of defect types and production parameters consisting of more than 100,000 samples, our study further includes information from real-time sensor data, imaging technologies, and historic production records. The results are that optimized machine learning models significantly improve defect detection compared to traditional methods. Take, for instance, the CNNs, which run at 98% and 96% accuracy in detecting packaging anomaly detection and bottle filling inconsistency, respectively, by fine-tuning the model with real-time imaging data, through which there was a reduction in false positives of about 30%. The optimized SVM model on detecting formulation defects gave 94% in viscosity variation detection and color variation. These values of performance metrics correspond to a giant leap in defect detection accuracy compared to the usual 80% level achieved up to now by rule-based systems. Moreover, this optimization with models can hasten defect characterization, allowing for detection time to be below 15 seconds from an average of 3 minutes using manual inspections with real-time processing of data. With this, the reduction in time will be combined with a 25% reduction in production downtime because of proactive defect identification, which can save millions annually in recall and rework costs. Integrating real-time machine learning-driven monitoring drives predictive maintenance and corrective measures for a 20% improvement in overall production efficiency. Therefore, the optimization of machine learning algorithms in defect characterization optimum scalability and efficiency for liquid detergent companies gives improved operational performance to higher levels of product quality. In general, this method could be conducted in several industries within the Fast moving consumer Goods industry, which would lead to an improved quality control process.

Keywords: liquid detergent manufacturing, defect detection, machine learning, support vector machines, convolutional neural networks, defect characterization, predictive maintenance, quality control, fast-moving consumer goods

Procedia PDF Downloads 20

Authors: Antoni Ivanov, Nikolay Dandanov, Nicole Christoff, Vladimir Poulkov

Abstract:

Spectrum underutilization has made cognitive radio a promising technology both for current and future telecommunications. This is due to the ability to exploit the unused spectrum in the bands dedicated to other wireless communication systems, and thus, increase their occupancy. The essential function, which allows the cognitive radio device to perceive the occupancy of the spectrum, is spectrum sensing. In this paper, the performance of modern adaptations of the four most widely used spectrum sensing techniques namely, energy detection (ED), cyclostationary feature detection (CSFD), matched filter (MF) and eigenvalues-based detection (EBD) is compared. The implementation has been accomplished through the PlutoSDR hardware platform and the GNU Radio software package in very low Signal-to-Noise Ratio (SNR) conditions. The optimal detection performance of the examined methods in a realistic implementation-oriented model is found for the common relevant parameters (number of observed samples, sensing time and required probability of false alarm).

Keywords: cognitive radio, dynamic spectrum access, GNU Radio, spectrum sensing

Procedia PDF Downloads 245

Authors: Xinwen Zhu, Xingguang Li, Sun Yi

Abstract:

Crack is one of the most common damages in buildings, bridges, roads and so on, which may pose safety hazards. However, cracks frequently happen in structures of various materials. Traditional methods of manual detection and measurement, which are known as subjective, time-consuming, and labor-intensive, are gradually unable to meet the needs of modern development. In addition, crack detection and measurement need be safe considering space limitations and danger. Intelligent crack detection has become necessary research. In this paper, an efficient method for crack detection and quantification using a 3D sensor, LiDAR, and depth camera is proposed. This method works even in a dark environment, which is usual in real-world applications. The LiDAR rapidly spins to scan the surrounding environment and discover cracks through lasers thousands of times per second, providing a rich, 3D point cloud in real-time. The LiDAR provides quite accurate depth information. The precision of the distance of each point can be determined within around  ±3 cm accuracy, and not only it is good for getting a precise distance, but it also allows us to see far of over 100m going with the top range models. But the accuracy is still large for some high precision structures of material. To make the depth of crack is much more accurate, the depth camera is in need. The cracks are scanned by the depth camera at the same time. Finally, all data from LiDAR and Depth cameras are analyzed, and the size of the cracks can be quantified successfully. The comparison shows that the minimum and mean absolute percentage error between measured and calculated width are about 2.22% and 6.27%, respectively. The experiments and results are presented in this paper.

Keywords: LiDAR, depth camera, real-time, detection and measurement

Procedia PDF Downloads 224

Authors: Abdullah Bal

Abstract:

This paper presents a one-class classification (OCC) technique based on Fukunaga-Koontz Transform (FKT) for binary classification problems. The FKT is originally a powerful tool to feature selection and ordering for two-class problems. To utilize the standard FKT for data domain description problem (i.e., one-class classification), in this paper, a set of non-class samples which exist outside of positive class (target class) describing boundary formed with limited training data has been constructed synthetically. The tunnel-like decision boundary around upper and lower border of target class samples has been designed using statistical properties of feature vectors belonging to the training data. To capture higher order of statistics of data and increase discrimination ability, the proposed method, termed one-class FKT (OC-FKT), has been extended to its nonlinear version via kernel machines and referred as OC-KFKT for short. Multiple kernel learning (MKL) is a favorable family of machine learning such that tries to find an optimal combination of a set of sub-kernels to achieve a better result. However, the discriminative ability of some of the base kernels may be low and the OC-KFKT designed by this type of kernels leads to unsatisfactory classification performance. To address this problem, the quality of sub-kernels should be evaluated, and the weak kernels must be discarded before the final decision making process. MKL/OC-FKT and selective MKL/OC-FKT frameworks have been designed stimulated by ensemble learning (EL) to weight and then select the sub-classifiers using the discriminability and diversities measured by eigenvalue ratios. The eigenvalue ratios have been assessed based on their regions on the FKT subspaces. The comparative experiments, performed on various low and high dimensional data, against state-of-the-art algorithms confirm the effectiveness of our techniques, especially in case of small sample size (SSS) conditions.

Keywords: ensemble methods, fukunaga-koontz transform, kernel-based methods, multiple kernel learning, one-class classification

Procedia PDF Downloads 21

Authors: Steffy Maria Joseph, Abdu Rahiman V, Abdul Hameed K. M.

Abstract:

Online handwritten character recognition is a challenging field in Artificial Intelligence. The classification success rate of current techniques decreases when the dataset involves similarity and complexity in stroke styles, number of strokes and stroke characteristics variations. Malayalam is a complex south indian language spoken by about 35 million people especially in Kerala and Lakshadweep islands. In this paper, we consider the significant feature extraction for the similar stroke styles of Malayalam. This extracted feature set are suitable for the recognition of other handwritten south indian languages like Tamil, Telugu and Kannada. A classification scheme based on support vector machines (SVM) is proposed to improve the accuracy in classification and recognition of online malayalam handwritten characters. SVM Classifiers are the best for real world applications. The contribution of various features towards the accuracy in recognition is analysed. Performance for different kernels of SVM are also studied. A graphical user interface has developed for reading and displaying the character. Different writing styles are taken for each of the 44 alphabets. Various features are extracted and used for classification after the preprocessing of input data samples. Highest recognition accuracy of 97% is obtained experimentally at the best feature combination with polynomial kernel in SVM.

Keywords: SVM, matlab, malayalam, South Indian scripts, onlinehandwritten character recognition

Procedia PDF Downloads 574

Authors: Kay Thinzar Phu, Lwin Lwin Oo

Abstract:

In an intelligent transport system and advanced driver assistance system, the developing of real-time traffic sign detection and recognition (TSDR) system plays an important part in recent research field. There are many challenges for developing real-time TSDR system due to motion artifacts, variable lighting and weather conditions and situations of traffic signs. Researchers have already proposed various methods to minimize the challenges problem. The aim of the proposed research is to develop an efficient and effective TSDR in real time. This system proposes an adaptive thresholding method based on RGB color for traffic signs detection and new features for traffic signs recognition. In this system, the RGB color thresholding is used to detect the blue and yellow color traffic signs regions. The system performs the shape identify to decide whether the output candidate region is traffic sign or not. Lastly, new features such as termination points, bifurcation points, and 90’ angles are extracted from validated image. This system uses Myanmar Traffic Sign dataset.

Keywords: adaptive thresholding based on RGB color, blue color detection, feature extraction, yellow color detection

Procedia PDF Downloads 313

Authors: Hyun-Woo Cho

Abstract:

Earlier detection of incipient abnormal operations in terms of plant-wide process management is quite necessary in order to improve product quality and process safety. And generating warning signals or alarms for operating personnel plays an important role in process automation and intelligent plant health monitoring. Various methodologies have been developed and utilized in this area such as expert systems, mathematical model-based approaches, multivariate statistical approaches, and so on. This work presents a nonlinear empirical monitoring methodology based on the real-time analysis of massive process data. Unfortunately, the big data includes measurement noises and unwanted variations unrelated to true process behavior. Thus the elimination of such unnecessary patterns of the data is executed in data processing step to enhance detection speed and accuracy. The performance of the methodology was demonstrated using simulated process data. The case study showed that the detection speed and performance was improved significantly irrespective of the size and the location of abnormal events.

Keywords: detection, monitoring, process data, noise

Procedia PDF Downloads 252

Authors: Shelby Simpson, William Stanley, Namir Naba, Xiaodi Wang

Abstract:

Both t-SNE and UMAP are brand new state of art tools to predominantly preserve the local structure that is to group neighboring data points together, which indeed provides a very informative visualization of heterogeneity in our data. In this research, we develop a t-SNE and UMAP base neural network image classification algorithm to embed the original dataset to a corresponding low dimensional dataset as a preprocessing step, then use this embedded database as input to our specially designed neural network classifier for image classification. We use the fashion MNIST data set, which is a labeled data set of images of clothing objects in our experiments. t-SNE and UMAP are used for dimensionality reduction of the data set and thus produce low dimensional embeddings. Furthermore, we use the embeddings from t-SNE and UMAP to feed into two neural networks. The accuracy of the models from the two neural networks is then compared to a dense neural network that does not use embedding as an input to show which model can classify the images of clothing objects more accurately.

Keywords: t-SNE, UMAP, fashion MNIST, neural networks

Procedia PDF Downloads 198

Authors: Deeman Yousif Mahmood, Mohammed Abdullah Hussein

Abstract:

The goal of a network-based intrusion detection system is to classify activities of network traffics into two major categories: normal and attack (intrusive) activities. Nowadays, data mining and machine learning plays an important role in many sciences; including intrusion detection system (IDS) using both supervised and unsupervised techniques. However, one of the essential steps of data mining is feature selection that helps in improving the efficiency, performance and prediction rate of proposed approach. This paper applies unsupervised K-means clustering algorithm with information gain (IG) for feature selection and reduction to build a network intrusion detection system. For our experimental analysis, we have used the new NSL-KDD dataset, which is a modified dataset for KDDCup 1999 intrusion detection benchmark dataset. With a split of 60.0% for the training set and the remainder for the testing set, a 2 class classifications have been implemented (Normal, Attack). Weka framework which is a java based open source software consists of a collection of machine learning algorithms for data mining tasks has been used in the testing process. The experimental results show that the proposed approach is very accurate with low false positive rate and high true positive rate and it takes less learning time in comparison with using the full features of the dataset with the same algorithm.

Keywords: information gain (IG), intrusion detection system (IDS), k-means clustering, Weka

Procedia PDF Downloads 296

Authors: P. Behera, K. K. Singh, D. K. Saini, M. De

Abstract:

Implementation of 2D materials in the detection of bio analytes is highly advantageous in the field of sensing because of its high surface to volume ratio. We have designed our sensor array with different cationic two-dimensional MoS₂, where surface modification was achieved by cationic thiol ligands with different functionality. Green fluorescent protein (GFP) was chosen as signal transducers for its biocompatibility and anionic nature, which can bind to the cationic MoS₂ surface easily, followed by fluorescence quenching. The addition of bio-analyte to the sensor can decomplex the cationic MoS₂ and GFP conjugates, followed by the regeneration of GFP fluorescence. The fluorescence response pattern belongs to various analytes collected and transformed to linear discriminant analysis (LDA) for classification. At first, 15 different proteins having wide range of molecular weight and isoelectric points were successfully discriminated at 50 nM with detection limit of 1 nM. The sensor system was also executed in biofluids such as serum, where 10 different proteins at 2.5 μM were well separated. After successful discrimination of protein analytes, the sensor array was implemented for bacteria sensing. Six different bacteria were successfully classified at OD = 0.05 with a detection limit corresponding to OD = 0.005. The optimized sensor array was able to classify uropathogens from non-uropathogens in urine medium. Further, the technique was applied for discrimination of bacteria possessing resistance to different types and amounts of drugs. We found out the mechanism of sensing through optical and electrodynamic studies, which indicates the interaction between bacteria with the sensor system was mainly due to electrostatic force of interactions, but the separation of native bacteria from their drug resistant variant was due to Van der Waals forces. There are two ways bacteria can be detected, i.e., through bacterial cells and lysates. The bacterial lysates contain intracellular information and also safe to analysis as it does not contain live cells. Lysates of different drug resistant bacteria were patterned effectively from the native strain. From unknown sample analysis, we found that discrimination of bacterial cells is more sensitive than that of lysates. But the analyst can prefer bacterial lysates over live cells for safer analysis.

Keywords: array-based sensing, drug resistant bacteria, linear discriminant analysis, two-dimensional MoS₂

Procedia PDF Downloads 143

Authors: M. Kamel, A. Hoayek, M. Batton-Hubert

Abstract:

With the increase of network virtualization and the disparity of vendors, the continuous monitoring and detection of anomalies cannot rely on static rules. An advanced analytical methodology is needed to discriminate between ordinary events and unusual anomalies. In this paper, we focus on log data (textual data), which is a crucial source of information for network performance. Then, we introduce an algorithm used as a pipeline to help with the pretreatment of such data, group it into patterns, and dynamically label each pattern as an anomaly or not. Such tools will provide users and experts with continuous real-time logs monitoring capability to detect anomalies and failures in the underlying system that can affect performance. An application of real-world data illustrates the algorithm.

Keywords: logs, anomaly detection, ML, scoring, NLP

Procedia PDF Downloads 94

Authors: A. Shoiynbek, K. Kozhakhmet, P. Menezes, D. Kuanyshbay, D. Bayazitov

Abstract:

Speech emotion recognition has received increasing research interest all through current years. There was used emotional speech that was collected under controlled conditions in most research work. Actors imitating and artificially producing emotions in front of a microphone noted those records. There are four issues related to that approach, namely, (1) emotions are not natural, and it means that machines are learning to recognize fake emotions. (2) Emotions are very limited by quantity and poor in their variety of speaking. (3) There is language dependency on SER. (4) Consequently, each time when researchers want to start work with SER, they need to find a good emotional database on their language. In this paper, we propose the approach to create an automatic tool for speech emotion extraction based on facial emotion recognition and describe the sequence of actions of the proposed approach. One of the first objectives of the sequence of actions is a speech detection issue. The paper gives a detailed description of the speech detection model based on a fully connected deep neural network for Kazakh and Russian languages. Despite the high results in speech detection for Kazakh and Russian, the described process is suitable for any language. To illustrate the working capacity of the developed model, we have performed an analysis of speech detection and extraction from real tasks.

Keywords: deep neural networks, speech detection, speech emotion recognition, Mel-frequency cepstrum coefficients, collecting speech emotion corpus, collecting speech emotion dataset, Kazakh speech dataset

Procedia PDF Downloads 101

Authors: Majid Habib Khan, Jinwei Zhao, Xinhong Hei, Liu Jiedong, Rana Shahzad Noor, Muhammad Imran

Abstract:

Deep learning is increasingly employed in image classification, yet the scarcity and high cost of labeled data for training remain a challenge. Limited samples often lead to overfitting due to biased sample distribution. This paper introduces a dynamic distribution calibration method for few-shot learning. Initially, base and new class samples undergo normalization to mitigate disparate feature magnitudes. A pre-trained model then extracts feature vectors from both classes. The method dynamically selects distribution characteristics from base classes (both adjacent and remote) in the embedding space, using a threshold value approach for new class samples. Given the propensity of similar classes to share feature distributions like mean and variance, this research assumes a Gaussian distribution for feature vectors. Subsequently, distributional features of new class samples are calibrated using a corrected hyperparameter, derived from the distribution features of both adjacent and distant base classes. This calibration augments the new class sample set. The technique demonstrates significant improvements, with up to 4% accuracy gains in few-shot classification challenges, as evidenced by tests on miniImagenet and CUB datasets.

Keywords: deep learning, computer vision, image classification, few-shot learning, threshold

Procedia PDF Downloads 67