Hybrid Machine Learning Algorithms Research Articles

Predicting sewer structural condition using hybrid machine learning algorithms

ABSTRACT Predicting the structural condition of sewer pipes plays a vital role in the predictive maintenance of sewer pipes and renewal plans of many water utilities. This study explores the simultaneous utilization of physical and environmental features of sewer pipes in sewer structural condition prediction. Three (3) hybrid machine learning models which are the combination of Bagging (BG), Dagging (DG), and Rotation Forest (RotF) ensembles with a J48 Decision Tree (J48DT) based classifier were used to predict sewer pipe conditions in Ålesund city, Norway. The classification performance of the machine learning models was evaluated using the area under the receiver operating characteristic (AUC-ROC) and the area under the precision-recall (AUC-PRC) curves. The RotF-J48DT model had the highest (AUC-ROC = 0.857, AUC-PRC = 0.918) values, followed by the BG-J48DT, and the base classifier J48DT. The RotF-J48DT hybrid model should be considered when predicting the condition of sewer pipes in the study area.

Hybrid Human and Machine Learning Algorithms to Forecast the European Stock Market

This paper explores the power of news sentiment to predict financial returns, particularly the returns of a set of European stocks. Building on past decision support work going back to the Delphi method, this paper describes a text analysis expert weighting algorithm that aggregates the responses of both humans and algorithms by dynamically selecting the best answer according to previous performance. The proposed system is tested through an experiment in which ensembles of experts, crowds, and machines analyzed Thomson Reuters news stories and predicted the returns of the relevant stocks mentioned right after the stories appeared. In most cases, the expert weighting algorithm was better than or as good as the best algorithm or human. The algorithm’s capacity to dynamically select the best answers from humans and machines results in an evolving collective intelligence: the final decision is an aggregation of the best automated individual answers, some of which come from machines and some from humans. Additionally, this paper shows that the groups of humans, algorithms, and expert weighting algorithms have associated with them, particularly, news topics that these groups are good at making predictions from.

Riverine flood potential assessment using metaheuristic hybrid machine learning algorithms

AbstractThis study presents the performance of stand‐alone and novel hybrid models combining the feed‐forward neural network (FFNN) and extreme gradient boosting (XGB) with the genetic algorithm (GA) optimization to determine the riverine flood potential at a local spatial scale, which is represented by the Gidra river basin, Slovakia. Eleven flood factors and a robust flood inventory database, consisting of 10,000 flood and non‐flood locations, were used. Using the FFNN, XGB, GA‐FFNN and GA‐XGB models, 16.5%, 11.0%, 17.1%, and 12.3% of the studied basin, respectively, is characterized with high to very high riverine flood potential. The applied models resulted in very high accuracy, that is, AUC = 0.93 in case of the FFNN stand‐alone model and AUC = 0.96 in case of the XGB stand‐alone model. The GA algorithm was able to raise the value of AUC for the hybrid GA‐FFNN and GA‐XGB models to 0.94 and 0.97, respectively. The results of this study can be useful, especially, for the identification of the areas with the highest potential for riverine floods within the next updating of the Preliminary Flood Risk Assessment, which is being carried out based on the EU Floods Directive.

An Ensemble Model for Forest Fire Occurrence Mapping in China

Assessing and predicting forest fires has long been an arduous task. Nowadays, the rapid advancement of artificial intelligence and machine learning technologies have provided a novel solution to forest fire occurrence assessment and prediction. In this research, we developed a novel hybrid machine-learning-technique algorithm to improve forest fire prediction based on random forest (RF), gradient-boosting decision tree (GBDT), support vector machine (SVM), and other machine learning models. The dataset we employed was satellite fire point data from 2010 to 2018 from the Chinese Department of Fire Prevention. The efficacy and performance of our methods were examined by validating the model fit and predictive capability. The results showed that the ensemble model LR (logistic regression)-RF-SVM-GBDT outperformed the single RFSVMGBDT model and the LR-RF-GBDT integrated framework, displaying higher accuracy and greater robustness. We believe that our newly developed hybrid machine-learning algorithm has the potential to improve the accuracy of predicting forest fire occurrences, thus enabling more efficient firefighting efforts and saving time and resources.

Prediction of permeability from well logs using a new hybrid machine learning algorithm

Permeability is a measure of fluid transmissibility in the rock and is a crucial concept in the evaluation of formations and the production of hydrocarbon from the reservoirs. Various techniques such as intelligent methods have been introduced to estimate the permeability from other petrophysical features. The efficiency and convergence issues associated with artificial neural networks have motivated researchers to use hybrid techniques for the optimization of the networks, where the artificial neural network is combined with heuristic algorithms. This research combines Social Ski-Driver (SSD) algorithm with the multilayer perception (MLP) neural network and presents a new hybrid algorithm to predict the value of rock permeability. The performance of this novel technique is compared with two previously used hybrid methods (Genetic Algorithm-MLP and Particle Swarm Optimization-MLP) to examine the effectiveness of these hybrid methods in predicting the permeability of the rock. The results indicate that the hybrid models can predict rock permeability with excellent accuracy. MLP-SSD method yields the highest coefficient of determination (0.9928) among all other methods in predicting the permeability values of the test data set, followed by MLP-PSO and MLP-GA, respectively. However, the MLP-GA converged faster than the other two methods and is computationally less expensive.

An automated software failure prediction technique using hybrid machine learning algorithms

An automated software failure prediction technique using hybrid machine learning algorithms

Using the artificial bee colony technique to optimize machine learning algorithms in estimating the mature weight of camels

This paper aims to predict male and female camels' mature weight (MW) through various morphological traits using hybrid machine learning (ML) algorithms. For this aim, biometrical measurements such as birth weight (BW), length of face (FL), length of the neck (NL), a girth of the heart (HG), body length (BL), withers height (WH), and hind leg length (HLL) were used to estimate the mature weight for eight camel breeds of Pakistan. In this study, multivariate adaptive regression splines (MARS), random forest (RF), and support vector machine (SVM) were applied to develop prediction models. Furthermore, the artificial bee colony (ABC) algorithm is employed to optimize ML models' internal parameters and improve prediction accuracy. The predictive performance of ML and hybrid models was evaluated on a testing dataset using goodness-of-fit measures such as mean absolute deviation (MAD), mean absolute percentage error (MAPE), coefficient of determination (R2), and root mean square error (RMSE). The results of the study revealed the ABC-SVM model was the best predictive model. The experimental results of this study showed that the proposed ABC-SVM method could effectively improve the accuracy for MW prediction of camels, thus having a research and practical value.

Enhancing flood susceptibility modeling using multi-temporal SAR images, CHIRPS data, and hybrid machine learning algorithms

Flood susceptibility maps are useful tool for planners and emergency management professionals in the early warning and mitigation stages of floods. In this study, Sentinel-1 dB radar images, which provide Synthetic-Aperture Radar (SAR) data were used to delineate flooded and non-flooded locations. 12 input parameters, including elevation, lithology, drainage density, rainfall, Normalized Difference Vegetation Index (NDVI), curvature, ground slope, Stream Power Index (SPI), Topographic Wetness Index (TWI), soil, Land Use Land Cover (LULC), and distance from the river, were selected for model development. The importance of each input parameter on flood occurrences was assessed via the Mutual Information (MI) technique. Several machine learning models, including Radial Basis Function (RBF), and three hybrid models of Bagging (BA-RBF), Random Committee (RC-RBF), and Random Subspace (RSS-RBF), were developed to delineate flood susceptibility areas at Goorganrood watershed, Iran. The performance of each model was evaluated using several error indicators, including correlation coefficient (r), Nash Sutcliffe Efficiency (NSE), Mean Absolute Error (MSE), Root Mean Square Error (RMSE), and the Area Under the Receiver Operating Characteristic (ROC) Curve (AUC). The results showed that the hybrid techniques enhanced the modeling performance of the standalone model, and generally, all hybrid models are more accurate than the standalone model. Although all developed models have performed well, RC-RBF outperforms all of them (AUC = 0.997), followed by BA-RBF (AUC = 0.996), RSS-RBF (AUC = 0.992), and RBF (AUC = 0.975). Generally, about 12 % of the study area has high and very high susceptibility to future flood occurrences.

A Hybrid Machine Learning Algorithm for Detection of Simulated Expiratory Markers of Diabetic Patients Based on Gas Sensor Array

The method for breath detection using gas sensor array is gaining popularity. It was found that acetone can be used as breath marker in diabetic patients. In this article, seven metal oxide gas sensors were used to collect acetone and ethanol gas, which are used to simulate the exhaled breath of diabetic patients to obtain multidimensional response data. The kernel principal component analysis (KPCA) algorithm is used to extract the characteristics from the data collected by the sensor array. The kind of gases is qualitatively identified using the adaptive boosting (AdaBoost) algorithm, and the grid search (GS) method is used to automatically optimize the parameters of AdaBoost algorithm. The quantitative analysis of gas concentration is performed using multivariate relevance vector machine (MVRVM) and it is also trained using the gas sensor array drift dataset at different concentrations from the University of California (UCI) database. The experimental results show that the accuracy of the algorithm in the qualitative identification of acetone and ethanol gas reaches 99.722%, and the root-mean-square errors (RMSE) for quantification of acetone and ethanol gases are 0.027 and 0.030, respectively. The algorithm is used for qualitative identification on the gas sensor array drift dataset at different concentrations with an accuracy of 94.55%, and the RMSE for quantification of acetone and ethanol gases are 11.59 and 8.72, respectively.

InSAR time-series analysis and susceptibility mapping for land subsidence in Semarang, Indonesia using convolutional neural network and support vector regression

Global sea-level rise due to climate change is a critical problem for coastal cities. One of the coastal cities in Indonesia, Semarang, is in danger of being submerged by seawater due to flooding, which is exacerbated by land subsidence. Time-series analysis based on persistent scatterer (PS) interferometry usually suffers from sparse measurement points (MP), and thus does not completely cover the study area with low coherence. In this study, we utilized improved combined scatterers interferometry with optimized point scatterers (ICOPS) to increase the MP density. We also compared the optimization of the ICOPS method based on support vector regression (SVR) with convolutional neural network (CNN) algorithms. The ICOPS-CNN showed better model performance and MP density, and was less time-consuming. Further land subsidence analysis was conducted in Semarang using susceptibility mapping. The land subsidence susceptibility map from the standalone deep learning (DL) algorithm (CNN = 0.833) outperformed the standalone machine learning (ML) algorithm (SVR = 0.820). The hybrid DL algorithm with grey wolf optimizer (GWO) (CNN-GWO = 0.907) exhibited the highest accuracy, followed by the hybrid ML algorithm (SVR-GWO = 0.843). The accuracy was consistent with the comparison between the susceptibility maps and the ICOPS measurements, resulting in a susceptibility map based on CNN-GWO with the highest accuracy (0.889). Land subsidence monitoring using ICOPS based on CNN and susceptibility mapping based on a hybrid DL algorithm (CNN-GWO) can be used by local governments to improve urban development planning in Semarang and its surrounding areas.

Automated Early Phase Breast Cancer Detection using Hybrid Machine Learning Algorithms

Breast cancer is the most common cancer among women. It occurs when few breast cells begin to grow abnormally. The national average for 2022 is 100.4 cases per 1,00,000 people, with a large number of women being diagnosed with breast cancer. The objective is to design a prediction system that can predict breast cancer at early stages using a set of attributes that have been selected from a critical dataset. The Wisconsin Kaggle dataset is used for this experiment. The goal of this work is to predict breast cancer utilizing hybrid machine learning methodologies, such as SVM and PCA. ML algorithms that could help to predict cancer, as the early detection of this disease would help to slow down the progression of other diseases. In our paper, we are implementing Hybrid algorithms like PCA and SVM and optimizing SVM with k-fold cross-validation for predicting Breast cancer at early stages with high accuracy. The goal is to raise the fraction of early-stage breast cancer detection and to reduce mistake rates with maximum precision, which are sustainable.

Robust Framework for Malevolent URL Detection using Hybrid Supervised Learning

The In this Internet Era, most of the people are using world-wide-web (www) based websites extensively for accomplishing their daily activities including E-commerce. As per latest report, total of 1.5 billion URLs are accessed every day and it is increasing every second. From another perspective, malicious URLs are employed in URL phishing- which in turn steal customer credentials and thereby lead to loss of billions of dollars. Due to these aspects, malevolent URL detection is gaining huge attention in the literature in the past few years. However, the state-of-the art literature employ only popular ML techniques and small datasets with less diversity for detecting malicious URLs. Further, the existing malicious URL detection methods are less focusing on bottleneck issues such as Overfitting and hyper-parameter tuning of ML models, which play a significant role in deciding the prediction accuracy of the proposed detection framework. In order to tackle these issues, this research article proposes a robust Malevolent URL Detection framework, which utilizes a hybrid machine learning algorithm, Support Vector Regression (SVR) and hyper parameters tuning strategies to enhance the prediction accuracy. The experimental results conducted on training and testing datasets clearly demonstrate the performance of the proposed detection framework in terms of PR and accuracy metrics respectively.

Digital twin application for attach detection and mitigation of PV-based smart systems using fast and accurate hybrid machine learning algorithm

Microgrids (MG) is originally designed to make Smart Grids more energy-efficient and reliable. The MG represents a complex cyber-physical system whose functioning is driven by the interaction of physical actions and computational elements, making it vulnerable to many forms of malicious cyber-attack. This study considers the effect of data integrity attack (DIA) on the performance of hybrid MGs, an important cyber threat to microgrids (MGs). Further, this paper develops a new process using sequential hypothesis testing (SHT) for detecting DIA on renewable energy resources and improving the security of information in hybrid MGs. By using a binary sample generated from the suggested approach, an analysis statistic is computed and afterward compared to 2 thresholds for deciding between the 3 options. On the measured energy production of renewable energy sources such as wind turbines, DIAs of various severity levels have been conducted for assessing the impact of DIAs on hybrid MG security. A standard IEEE test system has been used to evaluate the efficiency of the suggested process.

A Hybrid Machine Learning algorithm for Heart and Liver Disease Prediction Using Modified Particle Swarm Optimization with Support Vector Machine

Machine learning is now extensively applied in a variety of fields. Machine learning is employed like an efficient assistance mechanism in clinical diagnostics since vast amounts of data are readily available. Owing to heavy alcohol consumption, inhalation of contaminated gas, narcotics, food contamination, unhealthy life style people suffering from heart and liver disease has been significantly growing. Both heart and liver disease cause high mortality rate worldwide. It is critical to discover these diseases at an early stage in order to save people's lives. Incorporating machine learning classification algorithms into health-care organizations yields remarkable outcomes, allowing health-care practitioners to diagnose diseases more quickly and accurately. Machine learning techniques and tools aid in the extraction of useful information from datasets, resulting in more exact findings. In this study for heart and liver data classification, a hybrid model is created by combining support vector machine (SVM) approach and modified particle swarm optimization model. The data sets are collected from UCI machine learning repository. The results are calculated based on classification accuracy, error, correctness, recall as well as F1 score. The results obtained is compared with SVM, hybrid particle swarm optimization support vector machine algorithm (PSOSVM), hybrid Crazy particle swarm optimization support vector machine algorithm (CPSOSVM).

Mapping of Virtual Machines Using Machine Learning Algorithms for Detection of Faulty Nodes

A distributed system is characterized by a large number of nodes that are linked to a network and are mostly used for transaction processing. Large set of users are likely to communicate information over the network to the nodes, consistency and dependability remain a critical problem in the distributed environments. Independent failure of the component is one of the major problems in the distributed systems as it slowly impacts the performance of the other nodes in the system. The quality of service - QoS of a distributed network may be improved by a quick way of detecting problematic nodes. Sometime heavy nodes required high computation for transaction processing while idle nodes take low computation. In this paper, we proposed identification of straggler nodes in distributed environment with the help of hybrid machine learning algorithm. The work basically carried out to set up of large number of virtual machines and collect current log audits of each VM. According to the available parameters of audit files to each machine, algorithms decide that specific node is overheated or ideal condition. In expensive experimental analysis we demonstrate a accuracy of proposed hybrid machine learning algorithm. The proposed algorithm produces higher precision up to 4.5% than state-of-art methods. Key highlights of the VM mapping strategy were also investigated through a scrutiny of ongoing contracts. Main focus remains on machine learning (ML) to distinguish PM (Physical Machine) congestion, determining VMs from crowded PMs, and VM conditions as major exercises. This paper aims to review and characterize research on the planning and status of VMs that use ML using asset usage history. Energy productivity, VM migration, and quality of service were the main exhibition boundaries used to investigate cloud data center presentations.

Earlier diagnosis of breast cancer using e- health file with hybrid machine learning algorithms

Breast Cancer denotes the major disease which makes a high range of deaths worldwide. Earlier forecasting of breast cancer can help the breast cancer patients by improving their survival. Data mining methodologies and machine learning models are widely used in the health sector fields to offer effective diagnosis in various diseases. The main theme of this research work is to implement a hybrid approach of breast cancer predictions by using the SVM (Support vector Machine) and Naïve Bayes classifiers to predict the breast cancer using the E-Health file data of a patient. The proposed approach is aimed to offer the maximum accuracy than the existing approaches.

Comparison of Novel Hybrid and Benchmark Machine Learning Algorithms to Predict Groundwater Potentiality: Case of a Drought-Prone Region of Medjerda Basin, Northern Tunisia

Water scarcity is a severe problem in Tunisia, particularly in the northern region crossed by the Medjerda River, where groundwater is a conjoint water resource that is increasingly exploited. The aim of this study is to delineate the groundwater potential zones (GWPZs) in the Lower Valley of the Medjerda basin by using single benchmark machine learning models based on artificial neural network (ANN), random forest (RF), and support vector regression (SVR), and by developing a novel hybrid method, NB-RF-SVR, to reach the highest accuracy of groundwater potential prediction. Each model produced a spatial groundwater potential map (GPM) with the input of 26 groundwater-related factors (GRF) selected by the frequency ratio model and 70% of the transmissivity training data. The models’ effectiveness was assessed using the AUC-ROC curve, sensitivity, specificity, MAE, and RMSE metric indicators. The validation findings revealed that all the models performed successfully for the GWPZ mapping, where the AUC values for the ANN, RF, SVR, and NB-RF-SVR models were estimated as 71%, 79%, 87%, and 92%, respectively. The relative importance of the GWPZs revealed that land use followed by geology and elevation were the most important factors. Finally, these outcomes can provide valuable information for decision makers to effectively manage groundwater in water-stressed regions.

Modeling the mechanical properties of recycled aggregate concrete using hybrid machine learning algorithms

To explore the complicated functional relationship between key parameters such as the recycled aggregate properties, mix proportion and compressive strength of recycled aggregate concrete (RAC), a complete database involving 607 records from relevant published literature was built. Two standard algorithms (artificial neural network (ANN) and support vector regression (SVR)) and two optimized hybrid models (Particle Swarm Optimization based SVR (PSO-SVR) and grey Wolf optimizer based SVR (GWO-SVR)) were adopted. Furthermore, two interpretable algorithms (Partial Dependence Plot (PDP) and SHapley Additive exPlanations (SHAP)) were utilized to assess the global and local approaches independent of machine learning models, contributing towards decision-making rationales. Results indicated that the coefficient of determination (R2) of ANN, SVR, PSO-SVR and GWO-SVR were 0.7569, 0.5914, 0.8995 and 0.9056 respectively, showing that hybrid models outperformed the conventional models. However, GWO-SVR was the most problematic with overfitting when analyzing its three subsets. The two feature importance analyses revealed cement content, water content, natural fine aggregates, and water absorption as significant characteristics that affect mechanical performance.

Leveraging hybrid machine learning and data fusion for accurate mapping of malaria cases using meteorological variables in western India

We propose a hybrid machine learning algorithm (i.e.,P2CA−PSO−ANN) to model malaria outbreak in three districts (Barmer, Bikaner, and Jodhpur) of Rajasthan in the Western India. We have used different meteorological variables (i.e., relative humidity, temperature, and rainfall) as input features to predict malaria. We have also considered the combined impact of these variables through a linear data fusion. We then extract the uncorrelated information from the feature set by applying Probabilistic Principal Component Analysis (P2CA). We trained the fully connected feed-forward Artificial Neural Network (ANN) by optimising its hyperparameters iteratively through a bio-inspired optimisation algorithm (Particle Swarm Optimisation). We train and evaluate the performance of this algorithm using monthly meteorological variables from 2009 - 2012. This accurately predicts the malaria cases with the coefficient of correlation (R = 0.99), and Root Mean Square Error (RMSE = 1.76). Finally, we compare our model with different benchmark algorithms (Generalised Regression Neural Network (GRNN), Gaussian Process Regression (GPR), Support Vector Regression (SVR), Random Forest, and Radial Basis Neural Networks (RBNN)) in terms of accuracy. We observed the performance of hybrid machine learning model relatively high. This study can be used as an early warning intelligent system to predict the malaria outbreaks solely from meteorological data.

Multi-objective optimized high-strength concrete mix design using a hybrid machine learning and metaheuristic algorithm

Multi-objective optimized high-strength concrete mix design using a hybrid machine learning and metaheuristic algorithm