High Prediction Rate Research Articles

Genotypic characterization and antimicrobial susceptibility of human Campylobacter jejuni isolates in Southern Spain.

Campylobacter jejuni is the main cause of bacterial gastroenteritis and a public health problem worldwide. Little information is available on the genotypic characteristics of human C. jejuni in Spain. This study is based on an analysis of the resistome, virulome, and phylogenetic relationship, antibiogram prediction, and antimicrobial susceptibility of 114 human isolates of C. jejuni from a tertiary hospital in southern Spain from October 2020 to June 2023. The isolates were sequenced using Illumina technology, and a bioinformatic analysis was subsequently performed. The susceptibility of C. jejuni isolates to ciprofloxacin, tetracycline, and erythromycin was also tested. The resistance rates for each antibiotic were 90.3% for ciprofloxacin, 66.7% for tetracycline, and 0.88% for erythromycin. The fluoroquinolone resistance rate obtained is well above the European average (69.1%). CC-21 (n = 23), ST-572 (n = 13), and ST-6532 (n = 13) were the most prevalent clonal complexes (CCs) and sequence types (STs). In the virulome, the cadF, ciaB, and cdtABC genes were detected in all the isolates. A prevalence of 20.1% was obtained for the genes wlaN and cstIII, which are related to the pathogenesis of Guillain-Barré syndrome (GBS). The prevalence of the main antimicrobial resistance markers detected were CmeABC (92.1%), RE-cmeABC (7.9%), the T86I substitution in gyrA (88.9%), blaOXA-61 (72.6%), tet(O) (65.8%), and ant (6)-Ia (17.1%). High antibiogram prediction rates (>97%) were obtained, except for in the case of the erythromycin-resistant phenotype. This study contributes significantly to the knowledge of C. jejuni genomics for the prevention, treatment, and control of infections caused by this pathogen.IMPORTANCEDespite being the pathogen with the greatest number of gastroenteritis cases worldwide, Campylobacter jejuni remains a poorly studied microorganism. A sustained increase in fluoroquinolone resistance in human isolates is a problem when treating Campylobacter infections. The development of whole genome sequencing (WGS) techniques has allowed us to better understand the genotypic characteristics of this pathogen and relate them to antibiotic resistance phenotypes. These techniques complement the data obtained from the phenotypic analysis of C. jejuni isolates. The zoonotic transmission of C. jejuni through the consumption of contaminated poultry supports approaching the study of this pathogen through "One Health" approach. In addition, due to the limited information on the genomic characteristics of C. jejuni in Spain, this study provides important data and allows us to compare the results with those obtained in other countries.

Fast and affordable detection of PKU disease using iron (III) chloride-based solutions and porous PCL biosensors at higher prediction rates

Fast and affordable detection of PKU disease using iron (III) chloride-based solutions and porous PCL biosensors at higher prediction rates

An Oversampling Technique with Descriptive Statistics

Oversampling is often applied as a means to win a better knowledge model. Several oversampling methods based on synthetic instances have been suggested, and SMOTE is one of the representative oversampling methods that can generate synthetic instances of a minor class. Until now, the oversampled data has been used conventionally to train machine learning models without statistical analysis, so it is not certain that the machine learning models will be fine for unseen cases in the future. However, because such synthetic data is different from the original data, we may wonder how much it resembles the original data so that the oversampled data is worth using to train machine learning models. For this purpose, I conducted this study on a representative dataset called wine data in the UCI machine learning repository, which is one of the datasets that has been experimented with by many researchers in research for knowledge discovery models. I generated synthetic data iteratively using SMOTE, and I compared the synthetic data with the original data of wine to see if it was statistically reliable using a box plot and t-test. Moreover, since training a machine learning model by supplying more high-quality training instances increases the probability of obtaining a machine learning model with higher accuracy, it was also checked whether a better machine learning model of random forests can be obtained by generating much more synthetic data than the original data and using it for training the random forests. The results of the experiment showed that small-scale oversampling produced synthetic data with statistical characteristics that were statistically slightly different from the original data, but when the oversampling rate was relatively high, it was possible to generate data with statistical characteristics similar to the original data, in other words, after generating high-quality training data, and by using it to train the random forests, it was possible to generate random forests with higher accuracy than using the original data alone, from 97.75% to 100%. Therefore, by supplying additional statistically reliable synthetic data as a way of oversampling, it was possible to create a machine-learning model with a higher predictive rate.

A revised weight of evidence model for potential assessments of geothermal resources: a case study at western Sichuan Plateau, China

Efficient exploration of geothermal resources is the basis of exploitation and utilization of geothermal resources. In recent years, Geographic Information System (GIS) has been increasingly used for the exploration owing to its power ability to integrate and analyze multiple sources of data related to the formation of geothermal resources, such as geology, geophysics, and geochemistry. Correctly understanding the control effect of evidence factors on geothermal resources is the premise and basis of whether the prediction results of evidence weight model are accurate. Traditionally, the conventional weight of evidence model assume that each evidence factor exerts a uniform controlling effect on the formation and distribution of geothermal resources. However, recent research indicates significant variations in the controlling ability of factors such as faults and granites, influenced by factors like activity levels and crystalline ages. Yet, studies addressing this differential control are lacking. To address this gap, we propose a series of weight of evidence models using abundant geological, geophysical, and geothermal data from the western Sichuan plateau, a high-temperature geothermal hotspot in China. This study aims to investigate the impact of varying controlling abilities of evidence factors on the evaluation model, with faults and granites as a case. Performance metrics include prediction rate, success rate index, receiver operating characteristic curve (ROC) and prediction rate of geothermal well. The findings of this research reveal that the weight of evidence model developed through the methodology outlined in this study exhibits superior performance compared to the conventional weight of evidence model. This superiority is evidenced by higher prediction rates, success indices, prediction rate of geothermal wells, and larger AUC values of ROC. Among these models, the weight of evidence model considering both fault and granite classification have the best performance in model evaluation indicators, with a prediction rate of 22.528 and a success index of 0.015408 in the very high potential area. The prediction rate and success index of the high potential area are 3.656 and 0.0025, respectively, and the prediction rate and success index of the middle potential area are 1.649 and 0.001128, respectively, and the AUC value is 0.808, indicating that the model has good accuracy. In terms of geothermal well prediction, the total prediction rate of geothermal favorable areas based on fault and granite classification evidence weight model is as high as 47.0526. Therefore, when constructing the weight of evidence model, the influence of the difference control of evidence factors on the formation of geothermal resources should be fully considered. These results underscore the effectiveness of the proposed methodology in enhancing the predictive accuracy and reliability of geothermal resource assessment in this study. Based on the prediction results of the weight of evidence model considering both fault and granite classification, four favorable geothermal areas with abundant surface heat display are identified in this paper, namely Kangding, Litang, Batang and Ganzi-Dege. In addition, the relatively weak surface heat display areas such as Jiulong, Daofu, Luhuo and Derong also show high geothermal potential. Some attention should be paid to geothermal exploration in the future.

Gender Prediction Using Cone-Beam Computed Tomography Measurements from Foramen Incisivum: Application of Machine Learning Algorithms and Artificial Neural Networks

Introduction: This study aims to predict gender using parameters obtained from images of the foramen (for.) incisivum through cone-beam computed tomography (CBCT) and employing machine learning (ML) algorithms and artificial neural networks (ANN). Materials and Methods: This study was conducted on 162 individuals in total. Precise measurements were meticulously extracted, extending from the foramen incisivum to the arcus alveolaris maxillaris, through employment of CBCT. The ML and ANN models were meticulously devised, allocating 20% for rigorous testing and 80% for comprehensive training. Results: All parameters that are evaluated, except for the angle between foramen palatinum majus and foramen incisivum-spina nasalis posterior (GPFIFPNS-A), exhibited a significant gender difference. ANN and among the ML algorithms, logistic regression (LR), linear discriminant analysis (LDA), and random rorest (RF) demonstrated the highest accuracy (Acc) rate of 0.82. The Acc rates for other algorithms ranged from 0.76 to 0.79. In the models with the highest Acc rates, 14 out of 17 male individuals and 13 out of 16 female individuals in the test set were correctly predicted. Conclusion: LR, LDA, RF, and ANN yielded high gender prediction rates for the measured parameters, while decision tree, extra tree classifier, Gaussian Naive Bayes, quadratic discriminant analysis, and K-nearest neighbors algorithm methods provided lower predictions. We believe that the evaluation of measurements extending from foramen incisivum to arcus alveolaris maxillaris through CBCT scanning proves to be a valuable method in gender prediction.

Load Balance Forecasting Based on Hybrid Deep Neural Network

Load forecasting is the foundation of utility design, and it is a fundamental business problem in the utility industry. Load forecasting, mainly referring to forecasting electricity demand and energy, is being used throughout all segments of the electric power industry, including generation, transmission, distribution, and retail. In this paper, a long short-term memory network with a hybrid approach is improved with a dense algorithm and proposed for electricity load forecasting. A long short-term memory network is designed to effectively exhibit the dynamic behavior of load time series. The proposed model is tested for Panama study including historical data and weather variables. The prediction accuracy is validated by performance metrics, and the best of the metrics are attained when mean absolute error is 5.262, mean absolute percentage error 0.0000376, and root mean square error 18.243. The experimental results show a high prediction rate for load balance forecasting of electric power consumption.

Identifying patients with acute aortic dissection using an electrocardiogram with convolutional neural network

BackgroundThe potential of utilizing artificial intelligence with electrocardiography (ECG) for initial screening of aortic dissection (AD) is promising. However, achieving a high positive predictive rate (PPR) remains challenging. Methods and resultsThis retrospective analysis of a single-center, prospective cohort study (Shinken Database 2010–2017, N = 19,170) used digital 12-lead ECGs from initial patient visits. We assessed a convolutional neural network (CNN) model's performance for AD detection with eight-lead (I, II, and V1-6), single-lead, and double-lead (I, II) ECGs via five-fold cross-validation. The mean age was 63.5 ± 12.5 years for the AD group (n = 147) and 58.1 ± 15.7 years for the non-AD group (n = 19,023). The CNN model achieved an area under the curve (AUC) of 0.936 (standard deviation [SD]: 0.023) for AD detection with eight-lead ECGs. In the entire cohort, the PPR was 7 %, with 126 out of 147 AD cases correctly diagnosed (sensitivity 86 %). When applied to patients with D-dimer levels ≥1 μg/dL and a history of hypertension, the PPR increased to 35 %, with 113 AD cases correctly identified (sensitivity 86 %). The single V1 lead displayed the highest diagnostic performance (AUC: 0.933, SD: 0.03), with PPR improvement from 8 % to 38 % within the same population. ConclusionsOur CNN model using ECG data for AD detection achieved an over 30% PPR when applied to patients with elevated D-dimer levels and hypertension history while maintaining sensitivity. A similar level of performance was observed with a single-lead V1 ECG in the CNN model.

A Consistent Augmented Stacking Polynomial Optimized Tool (ASPOT) for Improving Security of Cloud-IoT Systems

The cloud computing transforms information technology by offering end users simulated, flexible resources on demand that require fewer resources and facilities giving them greater flexibility. These materials are delivered over the Internet using predefined networking protocols, regulations, and styles, and they are overseen by various management groups. There are flaws and vulnerabilities in the underlying technology and legacy protocols that could allow an attacker to get access. A recent assessment of the literature led to the conclusion that most intelligence algorithms have a number of complex issues, including a high false prediction rate, difficulty classifying threats, high processing costs, and system load. Hence, the proposed work aims to develop an innovative and lightweight Augmented Stacking Polynomial Optimized Tool (ASPOT) for strengthening the cloud-IoT system security against modern cyberattacks with a accuracy of 99%. The current study uses the lightweight Deep Augmented Preprocessing Model (DAPM) to clean and normalize the input cyber-attack dataset by executing transformation and normalization operations on it. Furthermore, the Binary Sand Cat Swarm Optimization (BSCSO) technique is utilized to identify the most significant and relevant features of the normalized dataset in an optimal manner. Moreover, the class of assaults is promptly and precisely identified from the given data by applying the Deep Stacking Polynomial Learning Network (DSPLN) technology. The effectiveness and results of the proposed ASPOT method are analysed with the use of current cyber-attack statistics and a variety of assessment metrics.

Detection of adulteration of cumin powder by front-face synchronous fluorescence spectroscopy: The influence of the natural variation of adulterants

For the authentication of powdered spices, the possible impact of the natural variations of adulterants on classification or prediction models is always ignored, and research in this area is still quite scarce. This study takes the application of front-face synchronous fluorescence spectroscopy (FFSFS) to the rapid and non-destructive authentication of cumin powder adulterated with ground peanut shells and maize flour as an example to show that how the natural variations of adulterants affect model prediction. Using three samples of each adulterant, two modes of sample set were designed for comparison. Mode 1 used the first adulterant for calibration and the other two for external validation. Mode 2 employed two adulterants in training and the third for external validation. The classification of adulterants by principal component analysis coupled with linear discriminant analysis (PCA–LDA) showed that mode 2 had a higher prediction rate (82%) in external validation than mode 1 (63%). For quantification, prediction models were built by partial least square (PLS) regression, and were validated by both cross- and external validation. The result of mode 2 was better than that of mode 1, with the determination coefficients of prediction (Rp2) greater than 0.93, the root mean square error of prediction (RMSEP) < 4.2% and residual predictive deviation (RPD) at least 2.7. This study demonstrates the necessity of the consideration of the natural variations of adulterants when constructing a robust model for adulteration detection.

Cell-Free Fetal DNA Screening Analysis in Korean Pregnant Women: Six Years of Experience and a Retrospective Study of 9327 Patients Analyzed from 2017 to 2022.

Cell-free DNA (cfDNA) screening for normal fetal aneuploidy has been widely adopted worldwide due to its convenience, non-invasiveness, and high positive predictive rate. We retrospectively evaluated 9327 Korean women with single pregnancies who underwent a non-invasive prenatal test (NIPT) to investigate how various factors such as maternal weight, age, and the method of conception affect the fetal fraction (FF). The average FF was 9.15 ± 3.31%, which decreased significantly as the maternal body mass index (BMI) increased (p < 0.001). The highly obese group showed a 'no-call' rate of 8.01%, which is higher than that of the normal weight group (0.33%). The FF was 8.74 ± 3.20% when mothers were in their 40s, and lower than that when in their 30s (9.23 ± 3.34, p < 0.001) and in the natural pregnancy group (9.31% ± 3.33). The FF of male fetuses was observed to be approximately 2.76% higher on average than that of female fetuses. As the gestational age increased, there was no significant increase in the fraction of fetuses up to 21 weeks compared to that at 10-12 weeks, and a significant increase was observed in the case of 21 weeks or more. The FFs in the NIPT high-risk result group compared to that in the low-risk group were not significantly different (p = 0.62). In conclusion, BMI was the factor most associated with the fetal fraction. Although the NIPT is a highly prevalent method in prenatal analysis, factors affecting the fetal fraction should be thoroughly analyzed to obtain more accurate results.

GWOKM: A novel hybrid optimization algorithm for geochemical anomaly detection based on Grey wolf optimizer and K-means clustering

Identifying the geochemical signatures of valuable mineral deposits using regional geochemical data from stream sediments is a challenging task due to the intricate characteristics of geological formations. Our team is currently investigating the potential of unsupervised clustering analysis (CA) and hybridization with the grey wolf optimizer (GWO) in developing multi-element geochemical models using stream sediment data. To cluster the geochemical data and uncover any unusual patterns, we opted to use the K-means (KM) algorithm due to its straightforward implementation, fast computation speed, and capacity to handle the large datasets. Despite its benefits, the KM method also has notable limitations, such as the random selection of cluster centroids. This can result in higher systematic uncertainty in unsupervised geochemical modeling and longer computation times. To mitigate this concern, we have introduced a new hybrid approach, grey wolf optimizer with K-means so-called the GWOKM algorithm to enhance the delineation of multi-elemental patterns in stream sediment geochemical data. This method incorporates the grey wolf optimization algorithm with KM to optimize the identification of both anomalies and backgrounds using factor analysis and sample catchment basin modeling techniques. This approach was utilized to detect anomalous multi-elemental geochemical patterns indicative of porphyry and skarn copper deposits in the Baft area, Kerman belt, Iran. Upon comparison of the geochemical models derived from KM and GWOKM clustering methods, the latter outperformed the former, as evidenced by its higher prediction rate. The outcomes affirm the efficacy of unsupervised KM clustering analysis (CA) as a means of breaking down geochemical anomaly-background populations. Moreover, the integration of clustering methods with optimization algorithms has proven to be successful for enhancing the credibility of mineralized areas, which could be advantageous in future exploration phases. Based on the results, the GWOKM approach generates more reliable and efficient geochemical anomaly targets.

Which pulse is it? Identifying archaeological legumes seeds by means of biometric measurements and geometric morphometrics

The taxonomic identification of archaeological seeds of certain cultivated pulses is challenging due to a combination of low interspecific morphological differences and large intraspecific variability. This study develops biometric and morphogeometric models that establish which species have more similarities with the doubtful archaeological pulse seeds. The species under study are: Lathyrus cicera/sativus (grass pea), Lens culinaris (lentil), Pisum sativum (pea), Vicia ervilia (bitter vetch), V. faba (broad bean), and V. sativa (vetch). The two models were trained on data from three different types of samples: contemporary uncharred seeds, contemporary charred seeds, and archaeological seeds. The last one consists of finds unearthed on Bronze and Iron Age sites throughout the Western Mediterranean and can be subdivided into two groups: clearly identified taxa and uncertain taxa. The biometric model resorts to the measurements of length, width and thickness of the three types of seeds so as to establish a discriminant linear model applicable to the uncertain archaeological seeds. The morphogeometric model is based on the shape of the three types of seeds gleaned from geometric morphometry. The uncertain archaeological seeds were then classified by means of a Linear Discriminant model based on shape descriptors. This study first assessed the accuracy of the two models stemming from observations of clearly defined contemporary and archaeological taxa. The results indicate that the models, in particular the morphogeometric, yield high predictive rates. These models therefore offer the possibility of re-identifying the taxa of doubtful archaeological seeds with a high degree of confidence and a minimal margin of error. The positive outcome of these models thus paves the way to more accurate determinations of archaeological legume seeds that heretofore have remained unidentified by traditional methods.

Machine-Learning-Based Diagnostics of Cardiac Sarcoidosis Using Multi-Chamber Wall Motion Analyses.

Hindered by its unspecific clinical and phenotypical presentation, cardiac sarcoidosis (CS) remains a challenging diagnosis. Utilizing cardiac magnetic resonance imaging (CMR), we acquired multi-chamber volumetrics and strain feature tracking for a support vector machine learning (SVM)-based diagnostic approach to CS. Forty-five CMR-negative (CMR(-), 56.5(53.0;63.0)years), eighteen CMR-positive (CMR(+), 64.0(57.8;67.0)years) sarcoidosis patients and forty-four controls (CTRL, 56.5(53.0;63.0)years)) underwent CMR examination. Cardiac parameters were processed using the classifiers of logistic regression, KNN(K-nearest-neighbor), DT (decision tree), RF (random forest), SVM, GBoost, XGBoost, Voting and feature selection. In a three-cluster analysis of CTRL versus vs. CMR(+) vs. CMR(-), RF and Voting classifier yielded the highest prediction rates (81.82%). The two-cluster analysis of CTRL vs. all sarcoidosis (All Sarc.) yielded high prediction rates with the classifiers logistic regression, RF and SVM (96.97%), and low prediction rates for the analysis of CMR(+) vs. CMR(-), which were augmented using feature selection with logistic regression (89.47%). Multi-chamber cardiac function and strain-based supervised machine learning provides a non-contrast approach to accurately differentiate between healthy individuals and sarcoidosis patients. Feature selection overcomes the algorithmically challenging discrimination between CMR(+) and CMR(-) patients, yielding high accuracy predictions. The study findings imply higher prevalence of cardiac involvement than previously anticipated, which may impact clinical disease management.

CFLCA: High Performance based Heart disease Prediction System using Fuzzy Learning with Neural Networks

Human Diseases are increasing rapidly in today’s generation mainly due to the life style of people like poor diet, lack of exercises, drugs and alcohol consumption etc. But the most spreading disease that is commonly around 80% of people death direct and indirectly heart disease basis. In future (approximately after 10 years) maximum number of people may expire cause of heart diseases. Due to these reasons, many of researchers providing enormous remedy, data analysis in various proposed technologies for diagnosing heart diseases with plenty of medical data which is related to heart disease. In field of Medicine regularly receives very wide range of medical data in the form of text, image, audio, video, signal pockets, etc. This database contains raw dataset which consist of inconsistent and redundant data. The health care system is no doubt very rich in aspect of storing data but at the same time very poor in fetching knowledge. Data mining (DM) methods can help in extracting a valuable knowledge by applying DM terminologies like clustering, regression, segmentation, classification etc. After the collection of data when the dataset becomes larger and more complex than data mining algorithms and clustering algorithms (D-Tree, Neural Networks, K-means, etc.) are used. To get accuracy and precision values improved with proposed method of Cognitive Fuzzy Learning based Clustering Algorithm (CFLCA) method. CFLCA methodology creates advanced meta indexing for n-dimensional unstructured data. The heart disease dataset used after data enrichment and feature engineering with UCI machine learning algorithm, attain high level accurate and prediction rate. Through this proposed CFLCA algorithm is having high accuracy, precision and recall values of data analysis for heart diseases detection.

Prediction for the Extent of Cancer based on Multiple Machine Learning Algorithms

Machine learning is a very important method for predicting the extent of cancer. In recent years, there has been considerable progress in detecting tiny features of the body, such as the specific surface area of cancer cells, the nuclear volume and perimeter of cancer cells. There is not development in measuring macro symptoms, such as Coughing of Blood, Chest Pain. Because in some remote and backward areas, detailed data are difficult to obtain, macro features need to be considered. In this paper, we use some relevant characteristics and environmental factors to predict the degree of cancer. During data preprocessing, some irrelevant contents are deleted, and the training set and the test set are divided. At the same time, some text data are digitized. Then, Naive Bayes, Decision Tree, Random Forest and Support Vector Machine were used in turn to make predictions and record their respective results. As a result, Naive Bayes, Decision Tree, Random Forest and Support Vector Machine achieved a fairly high prediction rate. More relevant features are obtained through the Person correlation coefficient and Feature Importance. On this basis, the prediction will not only maintain a high prediction rate, but also greatly reduce the memory consumption and training time.

Impact of LULC on debris flow using linear aggression model from Gilgit to Khunjerab with emphasis on urban sprawl.

In this research, the impact of land use and land cover (LULC) on debris flow was evaluated in the Gilgit to Khunjerab region. Two events have been done: (i) LULC stimulations for 2026 and 2030 using the MOLUSCE plugin and (ii) debris flow susceptibility mapping using linear aggression model. The evaluation of LULC on debris flow susceptibility is based on two scenarios: (i) existing (2010, 2014, 2018, 2022) LULC scenarios and (ii) stimulated (2026, 2030) LULC scenarios. The linear aggression model has 16 contributing factors to developing the debris flow susceptibility mapping. The main contributing components in debris flow susceptibility mapping are slope and LCCS. According to the linear aggressiveness model, debris flow susceptibility grows as the LULC changes, and the high susceptibility zones' share increases. For the current years 2010, 2014, 2018, and 2022, as well as the stimulated years 2026 and 2030, the model had high success rates (> 90.0%) and prediction rates (> 85.0%). The findings backed up prior research and suggested that the impact of LULC will grow in the future.

PhageTailFinder: A tool for phage tail module detection and annotation.

Decades of overconsumption of antimicrobials in the treatment and prevention of bacterial infections have resulted in the increasing emergence of drug-resistant bacteria, which poses a significant challenge to public health, driving the urgent need to find alternatives to conventional antibiotics. Bacteriophages are viruses infecting specific bacterial hosts, often destroying the infected bacterial hosts. Phages attach to and enter their potential hosts using their tail proteins, with the composition of the tail determining the range of potentially infected bacteria. To aid the exploitation of bacteriophages for therapeutic purposes, we developed the PhageTailFinder algorithm to predict tail-related proteins and identify the putative tail module in previously uncharacterized phages. The PhageTailFinder relies on a two-state hidden Markov model (HMM) to predict the probability of a given protein being tail-related. The process takes into account the natural modularity of phage tail-related proteins, rather than simply considering amino acid properties or secondary structures for each protein in isolation. The PhageTailFinder exhibited robust predictive power for phage tail proteins in novel phages due to this sequence-independent operation. The performance of the prediction model was evaluated in 13 extensively studied phages and a sample of 992 complete phages from the NCBI database. The algorithm achieved a high true-positive prediction rate (>80%) in over half (571) of the studied phages, and the ROC value was 0.877 using general models and 0.968 using corresponding morphologic models. It is notable that the median ROC value of 992 complete phages is more than 0.75 even for novel phages, indicating the high accuracy and specificity of the PhageTailFinder. When applied to a dataset containing 189,680 viral genomes derived from 11,810 bulk metagenomic human stool samples, the ROC value was 0.895. In addition, tail protein clusters could be identified for further studies by density-based spatial clustering of applications with the noise algorithm (DBSCAN). The developed PhageTailFinder tool can be accessed either as a web server (http://www.microbiome-bigdata.com/PHISDetector/index/tools/PhageTailFinder) or as a stand-alone program on a standard desktop computer (https://github.com/HIT-ImmunologyLab/PhageTailFinder).

Machine learning seizure prediction: one problematic but accepted practice

Objective. Epilepsy is one of the most common neurological disorders and can have a devastating effect on a person’s quality of life. As such, the search for markers which indicate an upcoming seizure is a critically important area of research which would allow either on-demand treatment or early warning for people suffering with these disorders. There is a growing body of work which uses machine learning methods to detect pre-seizure biomarkers from electroencephalography (EEG), however the high prediction rates published do not translate into the clinical setting. Our objective is to investigate a potential reason for this. Approach. We conduct an empirical study of a commonly used data labelling method for EEG seizure prediction which relies on labelling small windows of EEG data in temporal groups then selecting randomly from those windows to validate results. We investigate a confound for this approach for seizure prediction and demonstrate the ease at which it can be inadvertently learned by a machine learning system. Main results. We find that non-seizure signals can create decision surfaces for machine learning approaches which can result in false high prediction accuracy on validation datasets. We prove this by training an artificial neural network to learn fake seizures (fully decoupled from biology) in real EEG. Significance. The significance of our findings is that many existing works may be reporting results based on this confound and that future work should adhere to stricter requirements in mitigating this confound. The problematic, but commonly accepted approach in the literature for seizure prediction labelling is potentially preventing real advances in developing solutions for these sufferers. By adhering to the guidelines in this paper future work in machine learning seizure prediction is more likely to be clinically relevant.

Landslide susceptibility assessment on a large scale in the Podsljeme area, City of Zagreb (Croatia)

ABSTRACT The study presents a landslide susceptibility assessment on a large scale in the City of Zagreb (Croatia). The susceptibility analysis was performed using the Weight of Evidence model in the pilot area (21 km2) and applying the obtained weight values for each class of conditioning factors in the study area (130 km2). The input data were LiDAR-based landslide inventory and six conditioning factors derived from 5 m LiDAR DTM, 5 m SfM DEM, and geological and land-use maps. The validation of the susceptibility assessment for the study area was evaluated with a ROC curve, which showed a high prediction rate (AUC = 84.4%), similar to the prediction rate for the pilot area (AUC = 86.9%). Based on the results, it can be concluded that the proposed method for large-scale landslide susceptibility assessment, where susceptibility conditions are defined in smaller pilot areas, can be applied to larger research areas with similar geomorphological and geological conditions.

Improved Mask R-CNN Segmentation and Bayesian Interactive Adaboost CNN Classification for Breast Cancer Detection on Bach Dataset

Breast cancer is considered as the predominant type of cancer that affects more than ten percentage of the worldwide female population. Though microscopic evaluation remains to be a significant method for diagnosing, time and cost complexity seeks alternative and effective computer aided design for rapid and more accurate detection of the disease. As DL (Deep Learning) possess a significant contribution in accomplishing machine automation, this study intends to resolve existing problems with regard to lack of accuracy by proposing DL based algorithms. The study proposes Improved-Mask R CNN (I-MRCNN) method for segmentation. In this process, RPN (Region Proposal Network), predicts the objectless scores and object bound at every position. Here, (RoI Align) Region of interest Align is used for feature extraction as it is capable of resolving the conventional RoI pooling issues by attaining high accuracy for small objects and also eliminates quantization issues. Further, classification is performed using the proposed Bayesian Interactive Adaboost CNN classifier (B-IAB- CNN) that integrates the advantages of CNN, Bayesian and Adaboost classifier. The advantages of the three classifier enable optimum classification of the input Bach dataset that is confirmed through the results of performance analysis of the proposed system. Outcomes reveal that, average accuracy for segmentation is 96.32%, while, the classification accuracy is exposed to be 96%. As Timely prediction is significant, high prediction rate of the proposed system will assist the medical practitioners to detect breast cancer quickly which is the important practical implication from this study for diagnosing breast cancer.