Papers by Keyword: Support Vector Machine (SVM)

Abstract: In this paper, we propose a deep learning pipeline for real-time crop disease classification on mobile devices. Our system employs a custom Convolutional Neural Network (CNN) trained on publicly available crop disease datasets (Maize, Tomato, Potato, Rice). In addition, two transfer-learning models; ResNet-50 and MobileNet are used as fixed feature extractors, with their output features classified by a multi-class Support Vector Machine (SVM) with Radial Basis Function (RBF) kernel. We compare the models’ performance across all crop datasets and evaluate inference latency and model size. Experimental results show that the ResNet50-SVM hybrid attains near-perfect accuracy (≈100% for Maize, Tomato, Potato; 99.96% for Rice) on plant disease classification, far exceeding both the custom CNN and MobileNet-SVM approaches. The MobileNet-SVM pipeline is notably faster (≈23–66 ms per image) and compact (~8.7 MB) than ResNet50+SVM (≈108–192 ms, ~90 MB), making it well-suited for on-device deployment. The final model is converted to TensorFlow Lite for mobile inference; on a typical smartphone CPU it processes an input image in ~0.15–0.19s on average, enabling practical field use. These results demonstrate an efficient mobile AI solution for crop disease detection that balances accuracy with resource constraints. The proposed system can empower farmers with timely, in-field disease diagnosis, helping to mitigate yield losses and improve crop management through accessible AI-driven tools.

Abstract: Software-defined networking (SDN) infrastructures are facing a growing menace from cyber threats, which leave them vulnerable to distributed denial of service (DDoS) attacks. This study extends the earlier research [Detecting DDOS attacks in SDN Networks Using Machine Learning Techniques], which extensively investigated vulnerabilities in the SDN architecture for DDoS attacks. We develop and evaluate machine learning methods that are specifically tailored to protect software-defined networks (SDN) from such malicious attacks. The centralized and rigorous operational protocol of SDN has enabled us to develop a range of detecting methods. This study examines the efficacy of different machine learning algorithms, including XGBoost, Native Bayes, Logistic Regression, Support Vector Machines (SVM), Decision Trees, Random Forests, and K-Nearest Neighbors (KNN). The algorithm's computational efficiency and precision were evaluated using a dataset specifically created to simulate the intricate and unpredictable characteristics of network traffic. The experimental findings demonstrate that the XGBoost and Random Forest algorithms exhibit commendable performance in terms of both accuracy and speed. The precision varies between 99.26% and 77.0%, depending on the specific detection algorithm employed and the selected features. Consequently, these algorithms are well-suited for promptly dealing with and reducing potential hazards. XGBoost demonstrated exceptional versatility by maintaining accuracy across several testing scenarios while also reaching great processing efficiency. Utilizing machine learning has the potential to significantly enhance the security of SDN systems, as indicated by the results. This revelation has two significant ramifications: This study enhances our comprehension of efficient SDN DDoS mitigation strategies and sets a standard for the next research on integrating machine learning in security frameworks. We offer essential knowledge that aids in preserving vital network infrastructures in an ever more unpredictable digital landscape.

Abstract: Early fault diagnosis is a crucial element in maintaining the optimal operation of rotating machinery and avoiding sudden failure resulting in material and non-material losses. This research aims to select the salient features to diagnose the induction motor faults using an SVM model. The induction motor is simulated experiencing three fault scenarios: single fault, double faults, and multiple faults. These scenarios consist of stator fault, rotor fault, bearing fault, stator-bearing fault, stator-rotor fault, bearing-rotor fault, and stator-bearing-rotor fault. Vibration signals for each of these conditions are collected using an accelerometer sensor with a sampling frequency of 20 kHz. The study utilizes 12 statistical features, comprising 7-time time-domain features, namely mean, standard deviation, kurtosis, RMS, skewness, peak value, crest factor, and 5 frequency domain features, namely mean frequency, median frequency, spectral entropy, power spectral density, and spectral centroid. These features are selected using the ReliefF feature selection algorithm, and the selected features are then employed as classification parameters. The results indicate that the most relevant statistical features used for classification parameters are RMS, Standard Deviation, and Power Spectral Density. Meanwhile, the performance of the Support Vector Machine is excellent for three cases of the induction motor faults. The accuracies for single faults, double faults, and multiple faults are 99%, 100%, and 99% respectively.

Abstract: A support vector machine (SVM) is widely used for predicting the properties of fly ash blended concrete. However, the studies about the optimal design of fly ash blended concrete based on SVM are very limit. This study shows an SVM-based optimal design procedure of fly ash blended concrete. First, we built an SVM model and evaluated the compressive strength of fly ash blended concrete considering the effects of water to binder ratio, fly ash replacement ratio, and test ages. Second, we made parameter studies based on the SVM model. The parameter studies show that fly ash can improve the late age strength of concrete. This improvement is obvious for concrete with lower water to binder ratio. The optimal fly ash replacement ratio increases as the water to binder ratio decreases.

Abstract: Slope stability estimation is a complex engineering problem involving many factors. A hybrid model based on the combination of finite element software GEO-STUDIO and support vector machine (SVM) is proposed to address the problem. The study took a high slope of Jingjiang reach of Yangtze River as the research object. Several important parameters, including values of geometric and geotechnical properties of slope as well as rainfall and water level data were used to establish the finite element model for the high slope. Besides, the validity of the model was estimate using the measured data of pore water pressure. The slope stability coefficients were calculated in GEO-STUDIO environment. And the data were used as the input samples to train and test SVM model. Results show that the agreement achieved in pore water pressure between measurement and analysis using the finite element model can be considered very reasonable. And the slope stability coefficient results by SVM coincided well with that of finite element analysis. It suggests that the proposed model has the potential to be a useful tool for the prediction of slope stability coefficient considering the influence of rainfall and water level.

Abstract: The end effect of the local Mean Decomposition (LMD) causes serious distortion of the LMD decomposition results. And the most important factor of influence end effect is the extreme point and its distance, so the paper extracted the several factors, and composed of different sequences, using support vector machine (SVM) method respectively on the sets of data to predict, makes the original data can be extended. The research on the simulation signal and vibration signal shows that the method can effectively restrain the end effect of the decomposition.

Abstract: The oil and gas industry struggles to prevent formation of hydrates in pipeline by spending substantial amount of dollars. Hydrates are ice-like crystalline compounds that are composed of water and gas in which the gas molecules are trapped in water cavities. The hydrate formation is favorable at elevated pressure and reduced temperature and can be determined through experiment. However, the cost involved to determine early hydrate formation using experiment is driving researchers to seek for robust prediction methods using statistical and analytical methods. Main aim of the present study is to investigate applicability of radial basis function networks and support vector machines to hydrate formation conditions prediction. The data needed for modeling was taken from well-established literature. Performance of the models was assessed based on MSE, MAE, MAPE, MSPE, and Modified Pearson’s Correlation Coefficient. Data-based models enable the oil industry to predict the conditions leading to hydrate formation hence preventing clogging of the pipeline and high pressure buildup that could lead to sudden burst at the connections.

Abstract: This research presents a model for malware detection on mobile operating system based on analyzing the operation codes. The research processes are as follows: (1) achieving of both malicious and benign codes on android operating system, (2) extracting features based on the distribution of n-grams frequency where the parameter n = 3 is used, and (3) constructing a model for classification the malicious codes using the extracted features for both malicious and benign codes. In the experiment, 304 malicious codes and 553 benign codes were using to construct the model. The experiment shows that the model achieved more than 85.52% accuracy. For the sensitivity and specificity, the model achieved 71.26% and 90.52%, respectively.

Abstract: The friction welded joints made by GH4169 heat metal alloys are detected by U1traPAC system of the ultrasonic wave explore instrument. Aimed at the blemish signal characteristics, this article introduce Support Vector Machine (SVM) theory, which is based on statistical theory and structural risk minimization principle, to carry out multi-classification study of the detection signal. We decompose de-noising signals with wavelet packet transform, and extract energy eigenvalues according to "energy- defects". In accordance with designed "1-to-v" SVMs scheme, we respectively input normalized eigenvector to the SVM model to obtain the Forecast data. It is verificated that the limited existing data and information is well used by SVM and the signal is accurately been classificated. All of these verify that SVM has a strong generalization ability.

Abstract: Machines are the heart of most industries. By ensuring the health of machines, one could easily increase the company revenue and eliminates any safety threat related to machinery catastrophic failures. In condition monitoring (CM), questions often arise during decision making time whether the machine is still safe to run or not Traditional CM approach depends heavily on human interpretation of results whereby decision is made solely based on the individual experience and knowledge about the machines. The advent of artificial intelligence (AI) and automated ways for decision making in CM provides a more objective and unbiased approach for CM industry and has become a topic of interest in the recent years. This paper reviews the techniques used for automated decision making in CM with emphasis given on Dempster-Shafer (D-S) evident theory and other basic probability assignment (BPA) techniques such as support vector machine (SVM) and etc.