Automated Identification of Cotton and Viscose Fibers Using Image Analysis and Random Forest–Decision Tree Classification

Cross sections of fiber were treated with enzyme and the morphological changes were observed. The objective was obtaining information with respect to the degradation of fiber by enzyme. Viscose rayon, cuprammonium rayon, polynosic rayon, cotton and cellulose acetate were selected as the fiber samples and cellulase from Trichodelma viride was used as the enzyme. The fibers were embedded into epoxy resin and their cross sections were exposed by cutting with a microtome. Morphologies of the cross section of fiber were observed by scanning electron microscopy. As for cellulose acetate, morphological change was hardly obeserbed through the treatment with enzyme for 48 h, in comparison with untreated sample. The attack by enzyme caused crack in cotton fiber, but the same morphological change was not recognized in any fibers. Regarding the regenerated cellulosic fibers which were not chemically modified, such as viscose rayon, cuprammonium rayon and polynosic rayon, the changes were different from the other fibers, due to their higher-order structures.

Self-adaptive systems play a crucial role in the industrial world due to their ability to make corrections following environmental destabilization. In this paper we addressed the implementation of a machine learning approach which allows the detection of environmental hazards for the adaptation of information systems. We used internet connectivity data, which we collected from different networks, to train the model. We carried out these processes based on five machine learning algorithms, namely: XGBoost, Random Forest, SVM, Decision tree, K Neighbors. The results obtained by its algorithms in terms of test accuracy are 99.95% for XGBoost, Random Forest Decision tree, and 97.62% for SVM and 99.80% for K Neighbors. In terms of train accuracy are 100% for XGBoost, Random Forest Decision tree, and 97.74% for SVM and 99.84% for K Neighbors. In terms of F1-score are 99.95% for and 97.80% for K Neighbors. In terms of precision are 99.95% for XGBoost, Random Forest Decision tree, and 97.75% for SVM and 97.81% for K Neighbors. In terms of recall are 99.95% for XGBoost, Random Forest Decision tree, and 97.62% for SVM and 97.80% for K Neighbors. We also carried out a comparative analysis of the results obtained by each of the five algorithms, based on the following performance indicators: accuracy, F1-score, precision, Recall.

Long-lived and healthy lifestyles are the means for sustainable physical and mental wellbeing. Physical activity monitoring is one of the key aspects of maintaining a healthy lifestyle. It can be achieved with the help of widely used smartphones containing accelerometers sensors. The current study proposes the applicability of different machine learning methods for physical activity monitoring based on accelerometer sensor data and their classification. In this paper, Radial Basis Function network (RBF), Multilayer Perceptron (MLP), K-nearest neighbors classifier (KNN) and Random Forest decision tree (RFDT) have been employed for the classification task. Among the classification algorithms, RFDT and KNN recognize all activities very well (97.5% and 96.8%, respectively) with less confusion than other algorithms. Based on accuracy and error scores, it is found that RFDT with the lowest root-mean-square error (0.0758) produce better outcomes than other algorithms. This study observed that RFDT has an important role in classifying the physical activities with better accuracy within a short span of time. Further investigation will be carried out to find out the correlation of physical activity monitoring with different diseases.

As of today, it still remains a considerable challenge to identify the polyester/cotton fiber category in the textile industry. Based on shape coefficient analysis, a novel comprehensively-designed and appropriately-developed digital cross-section image processing method was proposed to improve the identification accuracy of polyester/cotton fiber. A self-developed microscope image acquisition system was established, so as to digitalize the cross section of fiber. Aimed at the gray scale enhancement, binarization and opening operation, a set of image preprocessing algorithms was developed. Initially, the edge detection algorithm was used to identify the external contour of the cross-section of fiber. Subsequently, the cross section of fiber was separated by watershed segmentation algorithm, and that, the false fiber was removed by area elimination algorithm. Finally, the shape coefficient was calculated to identify polyester/cotton fiber category. The experimental results indicated that the data and results obtained by the newly-designed digital analysis method which is based on shape coefficient in this paper can be used to identify the polyester/cotton fiber category, showing a broad application prospect in textile quality control, including the developments of concerned equipments, methods and standards.

Predictive mapping of seabed sediments based on multibeam bathymetric (BM), and backscatter (BS) data is effective for mapping the spatial distribution of the substrate. A robust modeling technique, the random forest decision tree (RFDT), was used to predict the seabed sediments in an area of the Joseph Bonaparte Gulf, Northern Australia, using the multibeam data and seabed sediment samples collected simultaneously. The results showed that: (1) Using multibeam bathymetry data in addition to multibeam backscatter data improves the prediction performance of the RFDT. In comparison to only multibeam backscatter data, the prediction performance achieved a ~10% improvement in sediment properties; it achieved a ~44.45% improvement of overall accuracy in sediment types, and a ~0.55 improvement in Kappa. (2) The underlying relationships between sediment properties and multibeam data show that there is an opposite non-linear correlation between sediment property-BS and sediment property-BM. For example, there is an obvious negative relationship between %mud-BS at incidence angles of 13° and 21°, but the relationship between %mud-BM is positive. As such, the RFDT is a useful and well-performing method in predicting the relationship between sediment properties and multibeam data and in predicting the distribution of sediment properties and types. However, the sediment prediction method in deep-water areas with high gravel content needs to be further evaluated.

Hierarchical pattern recognition in milking parameters predicts mastitis prevalence

We present a method for 3D measurement of fiber cross sectional morphology from handsheets. An automated procedure is used to acquire 3D datasets of fiber cross sectional images using an automated microtome and light microscopy. The fiber cross section geometry is extracted using digital image analysis. Simple sample preparation and highly automated image acquisition and image analysis are providing an efficient tool to analyze large samples. It is demonstrated that if fibers are tilted towards the image plane the images of fiber cross sections are always larger than the true fiber cross section geometry. In our analysis the tilting angles of the fibers to the image plane are measured. The resulting fiber cross sectional images are distorted to compensate the error due to fiber tilt, restoring the true fiber cross sectional shape. We use an approximated correction, the paper provides error estimates of the approximation. Measurement results for fiber wall thickness, fiber coarseness and fiber collapse are presented for one hardwood and one softwood pulp.

The lack of health professionals, quality health care, and accessible health centers in rural and isolated communities have resulted in high maternal and fetal mortality rates in the Philippines, especially since quality maternal and child healthcare services are concentrated in more developed urban areas. This study proposes applying the Telemedicine framework as a helping tool for doctors and health professionals. The implemented Telemedicine approach on prenatal care followed a set-up that included patient information input, a mobile application used for both data input and visualization, a cloud-based server for the database, and a machine learning system which analyzed data from the patient profile. With a dataset of 97 samples, four algorithms were implemented for the development of the machine learning system - Decision Tree, Random Forest Decision Tree, K-Nearest Neighbor, and Support Vector Machine. The performance of each algorithm was tested in terms of accuracy, precision, recall, and the F1 score or the weighted average of precision and recall. Based on these parameters, the most effective algorithm was the Random Forest Decision Tree with the highest train score (0.987) and test score (0.900). The results of this algorithm were visualized in an Android mobile application that displayed whether the patient was a positive or negative case with respect to the possibility of having a high-risk pregnancy.

Application of machine learning to the prediction of postoperative sepsis after appendectomy

The article is devoted to the current topic of analysis of mathematical models for countering cyber fraud in banks. This problem is due to the security risks growth in the banking system, which are formed by fraudsters' cyberattacks and cybercrimes implementation. Therefore, the priority task for cyberbanking security is the application of modern mathematical methods to analyse the sources of cyber attacks, identify threats and losses in the banking services market, identify cyber-attacks and assess the scenario of potential cyber risk, etc. The article analyses the most widespread types of cyber fraud: social engineering, phishing, stalking, farming, DoS attacks, online fraud, potentially unwanted programs, etc. The study also considered a model of cognitive computing and detection of suspicious transactions in banking cyber-physical systems based on quantum computing in BCPS for the post-quantum era. The advantages, disadvantages and results of the model are defined. Predictive modelling is proposed to detect fraud in real-time by analysing incoming bank transactions with payment cards. Within the framework of this method, such models are used for the classification of fraud detection as logistic regression, a decision tree, and a narrower technique - a random forest decision tree. The study also considered using the harmonic search algorithm in neural networks to improve fraud detection in the banking system. It is found that although this model has the advantage of learning ability based on past behaviour, there are difficulties in the long-term processing of many neural networks. The stages of model implementation are also given. In addition, the modelling of credit card fraud detection is based on using two types of models: supervised and unsupervised. Supervised models include logistic regression, K-nearest neighbours, and extreme gradient boosting. The one-class support vector model, restricted Boltzmann model, and generative-competitive network are considered among uncontrolled generative models.

MANETs are self-organizing network architectures of mobile nodes. Due to node mobility, wireless network topologies dynamically various over time. A novel link stability estimation technique called Hybridization of Brownboost Cluster and Random Forest Decision Tree with Optimized Route Selection (HBCRFDT-GORS) technique is introduced for increasing the reliable data delivery by eliminating the stale routes in MANET. Brown Boost technique is applied to find the route paths having the smaller number of hop counts to perform the data transmission. After that, the status of the mobile nodes in the selected route paths is determined based on the residual energy and signal strength. Then, a random forest decision tree is applied to correctly identify the stale routes by finding the link failure due to the selfish node and the corruptive node along the route path. Then the broken link is removed from the route path. After eliminating the stale route from the path, the HBCRFDT-GORS technique finds the alternative optimal route through the gradient free optimization. The proposed HBCRFDT-GORS technique performs stale route elimination and improves reliable data delivery from source to destination. Simulation is conducted on different performance metrics such as routing overhead, packet delivery ratio, packet drop rate, and delay with respect to the number of data packets. The Network simulation results indicate that the HBCRFDT-GORS technique is improving the data delivery and and minimizing the delay as well as reducing the packet losses when compared to the baseline approaches.

Maturity of cotton fibers refers to the extent of development of their secondary wall and is an important characteristic. To characterize and quantify cotton maturity, we have developed an imaging system based on our previous research on fiber cross- sectional shape analysis. In this paper, we focus on methods of characterizing cotton fiber maturity. We review the techniques for processing images of fiber cross sections, propose a new approach for describing cotton maturity and discuss other related char acteristics.

A joint and simultaneous prediction framework of weekday and weekend daily-activity travel pattern using conditional dependency networks

BackgroundRecent advances in MRI acquisitions and image analysis have increased the utility of neuroimaging in understanding disease-related changes. In this work, we aim to demonstrate increased sensitivity to disease progression as well as improved diagnostic accuracy in Amyotrophic lateral sclerosis (ALS) with multimodal MRI of the brain and cervical spinal cord.MethodsWe acquired diffusion MRI data from the brain and cervical cord, and T1 data from the brain, of 20 participants with ALS and 20 healthy control participants. Ten ALS and 14 control participants, and 11 ALS and 13 control participants were re-scanned at 6-month and 12-month follow-ups respectively. We estimated cross-sectional differences and longitudinal changes in diffusion metrics, cortical thickness, and fixel-based microstructure measures, i.e. fiber density and fiber cross-section.ResultsWe demonstrate improved disease diagnostic accuracy and sensitivity through multimodal analysis of brain and spinal cord metrics. The brain metrics also distinguished lower motor neuron-predominant ALS participants from control participants. Fiber density and cross-section provided the greatest sensitivity to longitudinal change. We demonstrate evidence of progression in a cohort of 11 participants with slowly progressive ALS, including in participants with very slow change in ALSFRS-R. More importantly, we demonstrate that longitudinal change is detectable at a six-month follow-up visit. We also report correlations between ALSFRS-R and the fiber density and cross-section metrics.ConclusionsOur findings suggest that multimodal MRI is useful in improving disease diagnosis, and fixel-based measures may serve as potential biomarkers of disease progression in ALS clinical trials.

PurposeThe face recognition problem has a long history and a significant practical perspective and one of the practical applications of the theory of pattern recognition, to automatically localize the face in the image and, if necessary, identify the person in the face. Interests in the procedures underlying the process of localization and individual’s recognition are quite significant in connection with the variety of their practical application in such areas as security systems, verification, forensic expertise, teleconferences, computer games, etc. This paper aims to recognize facial images efficiently. An averaged-feature based technique is proposed to reduce the dimensions of the multi-expression facial features. The classifier model is generated using a supervised learning algorithm called a back-propagation neural network (BPNN), implemented on a MatLab R2017. The recognition rate and accuracy of the proposed methodology is comparable with other methods such as the principle component analysis and linear discriminant analysis with the same data set. In total, 150 faces subjects are selected from the Olivetti Research Laboratory (ORL) data set, resulting 95.6 and 85 per cent recognition rate and accuracy, respectively, and 165 faces subjects from the Yale data set, resulting 95.5 and 84.4 per cent recognition rate and accuracy, respectively.Design/methodology/approachAveraged-feature based approach (dimension reduction) and BPNN (generate supervised classifier).FindingsThe recognition rate is 95.6 per cent and recognition accuracy is 85 per cent for the ORL data set, whereas the recognition rate is 95.5 per cent and recognition accuracy is 84.4 per cent for the Yale data set.Originality/valueAveraged-feature based method.