Application Of Machine Learning Approaches Research Articles

Improved Bst DNA Polymerase Variants Derived via a Machine Learning Approach.

The DNA polymerase I from Geobacillus stearothermophilus (also known as Bst DNAP) is widely used in isothermal amplification reactions, where its strand displacement ability is prized. More robust versions of this enzyme should be enabled for diagnostic applications, especially for carrying out higher temperature reactions that might proceed more quickly. To this end, we appended a short fusion domain from the actin-binding protein villin that improved both stability and purification of the enzyme. In parallel, we have developed a machine learning algorithm that assesses the relative fit of individual amino acids to their chemical microenvironments at any position in a protein and applied this algorithm to predict sequence substitutions in Bst DNAP. The top predicted variants had greatly improved thermotolerance (heating prior to assay), and upon combination, the mutations showed additive thermostability, with denaturation temperatures up to 2.5 °C higher than the parental enzyme. The increased thermostability of the enzyme allowed faster loop-mediated isothermal amplification assays to be carried out at 73 °C, where both Bst DNAP and its improved commercial counterpart Bst 2.0 are inactivated. Overall, this is one of the first examples of the application of machine learning approaches to the thermostabilization of an enzyme.

Explainable Machine learning on New Zealand strong motion for PGV and PGA

Estimating ground motion characteristics at various locations as a function of fault characteristics is useful for the proper damage assessment and risk mitigation strategies. This paper explores the application of machine learning approaches to predict peak ground acceleration (PGA) and peak ground velocity (PGV) using New Zealand’s strong motion data. Five machine learning algorithms, namely linear regression, kNN, SVM, Random Forest, and XGBoost, are used in this study. Using the New Zealand flat-file database, the geometric mean of the peak ground motion parameters is used as predictor variables in training the machine learning algorithms. The performance of the chosen algorithms and how they work on PGV and PGA are discussed. The best prediction for PGA is obtained using random forest but for PGV XGboost worked best. The relative importance of various features in the flat file is also presented for the best-performing machine learning algorithm. Although the magnitude of an earthquake is found to be most influential for PGV, rupture distance showed the highest impact for PGA. Finally, the predictions are also explained using SHApley Additive exPlanations (SHAP) for the overall dataset as well as on a sample by sample basis, for a few samples. Pairwise dependency of some features with the highest feature importance is also presented using SHAP.

Application of machine learning approaches to predict the 5-year survival status of patients with esophageal cancer.

BackgroundAccurate prognostic estimation for esophageal cancer (EC) patients plays an important role in the process of clinical decision-making. The objective of this study was to develop an effective model to predict the 5-year survival status of EC patients using machine learning (ML) algorithms.MethodsWe retrieved the information of patients diagnosed with EC between 2010 and 2015 from the Surveillance, Epidemiology, and End Results (SEER) Program, including 24 features. A total of 8 ML models were applied to the selected dataset to classify the EC patients in terms of 5-year survival status, including 3 newly developed gradient boosting models (GBM), XGBoost, CatBoost, and LightGBM, 2 commonly used tree-based models, gradient boosting decision trees (GBDT) and random forest (RF), and 3 other ML models, artificial neural networks (ANN), naive Bayes (NB), and support vector machines (SVM). A 5-fold cross-validation was used in model performance measurement.ResultsAfter excluding records with missing data, the final study population comprised 10,588 patients. Feature selection was conducted based on the χ2 test, however, the experiment results showed that the complete dataset provided better prediction of outcomes than the dataset with removal of non-significant features. Among the 8 models, XGBoost had the best performance [area under the receiver operating characteristic (ROC) curve (AUC): 0.852 for XGBoost, 0.849 for CatBoost, 0.850 for LightGBM, 0.846 for GBDT, 0.838 for RF, 0.844 for ANN, 0.833 for NB, and 0.789 for SVM]. The accuracy and logistic loss of XGBoost were 0.875 and 0.301, respectively, which were also the best performances. In the XGBoost model, the SHapley Additive exPlanations (SHAP) value was calculated and the result indicated that the four features: reason no cancer-directed surgery, Surg Prim Site, age, and stage group had the greatest impact on predicting the outcomes.ConclusionsThe XGBoost model and the complete dataset can be used to construct an accurate prognostic model for patients diagnosed with EC which may be applicable in clinical practice in the future.

Application of Machine Learning Approaches in Intrusion Detection System

The rapid development of technology reveals several safety concerns for making life more straightforward. The advance of the Internet over the years has increased the number of attacks on the Internet. The IDS is one supporting layer for data protection. Intrusion Detection Systems (IDS) offer a healthy market climate and prevent misgivings in the network. Recently, IDS has been used to recognize and distinguish safety risks using Machine Learning (ML). This paper proposed a comparative analysis of the different ML algorithms used in IDS and aimed to identify intrusions with SVM, J48, and Naive Bayes. Intrusion is also classified. Work with the KDD-CUP data set, and their performance has been checked with the WEKA software. A comparison of techniques such as J48, SVM, and Naïve Bayes showed that the accuracy of j48 is the higher one which was (99.96%).

A weakly supervised deep learning approach for label-free imaging flow-cytometry-based blood diagnostics

SummaryThe application of machine learning approaches to imaging flow cytometry (IFC) data has the potential to transform the diagnosis of hematological diseases. However, the need for manually labeled single-cell images for machine learning model training has severely limited its clinical application. To address this, we present iCellCnn, a weakly supervised deep learning approach for label-free IFC-based blood diagnostics. We demonstrate the capability of iCellCnn to achieve diagnosis of Sézary syndrome (SS) from patient samples on the basis of bright-field IFC images of T cells obtained after fluorescence-activated cell sorting of human peripheral blood mononuclear cell specimens. With a sample size of four healthy donors and five SS patients, iCellCnn achieved a 100% classification accuracy. As iCellCnn is not restricted to the diagnosis of SS, we expect such weakly supervised approaches to tap the diagnostic potential of IFC by providing automatic data-driven diagnosis of diseases with so-far unknown morphological manifestations.

Time period estimation of masonry infilled RC frames using machine learning techniques

The accurate estimation of the fundamental time period is critical for the error-free risk and reliability estimation of infrastructure systems. Although complex empirical models are available in the literature, this paper estimates the application of machine learning approaches for the time period estimation. Recently, a good database of masonry-infilled RC frames and their fundamental period exist in literature and preliminary approaches like artificial neural networks have been tried out on them. In this work, we use advanced machine learning algorithms based on bagging and boosting approaches, and the comparison of our results with those already published shows that these methods can outperform the existing ones. The contribution of each variable to the fundamental time period is explained locally and globally using Shapely Additive Explanations.

Predictive classification and understanding of weather impact on airport performance through machine learning

Efficient airport operations depend on appropriate actions and reactions to current constraints. Local weather events and their impact on airport performance may have network-wide effects. The classification of expected weather impacts enables efficient consideration in airport operations on a tactical level. We classify airport performance with recurrent and convolutional neural networks considering weather data. We are using London–Gatwick Airport to apply our developed approach. The weather data is derived from local meteorological reports and airport performance is derived from both flight plan data and reported delays. We show that the application of machine learning approaches is an appropriate method to quantify the correlation between decreased airport performance and the severity of local weather events. The developed models could achieve prediction accuracy higher than 90% for departure movements. We see our approach as one key element for a deeper understanding of interdependencies between local and network operations in the air transportation system.

The Application of Machine Learning Approaches on Real-Time Apartment Prices in the Tokyo Metropolitan Area

AbstractThe widely applied hedonic regression approach for the relationship between property prices and housing attributes is subject to assumptions and specifications of models as well as the availability and content of second-hand official data. In a cross-disciplinary spirit, this study employs machine learning techniques to examine hedonic apartment prices in the Tokyo Metropolitan Area of Japan based on online sales data extracted by web-parsing technology. With 14,579 apartment observations, two machine learning regressions—decision tree (DT) and random forest (RF)—are compared to conventional ordinary least squares regression (OLS) for hedonic modelling. Empirical results demonstrated that RF regressions led to the highest accuracy in model prediction performance, followed by DT and OLS. The comparison with results across models revealed that the housing features that have consistent influences on apartment prices tend to be those associated with living quality (including management funds, repair fund fees, floor size, located floor, total floor of the building, and location in Tokyo). Other commonly appreciated features, such as southward orientation or corner-lot location, did not demonstrate importance, possibly due to changes in residents’ preferences. In this big-data era, the adaptation of real-time data and machine learning approaches should add value to the variable selection process and model performance.

An empirical investigation into intelligent cost analysis in purchasing

PurposeCost transparency is of central importance to reach a consensus between supply chain partners. The purpose of this paper is to contribute to the instrument of cost analysis which supports the link between buyers and suppliers.Design/methodology/approachBased on a detailed literature review in the area of cost analysis and purchasing, intelligent decision support systems for cost estimation are identified. Subsequently, expert interviews are conducted to determine the application possibilities for managers. The application potential is derived from the synthesis of motivation, identified applications and challenges in the industry. Management recommendations are to be derived by bringing together scientific and practical approaches in the industry.FindingsOn the one hand, the results of this study show that machine learning (ML) is a complex technology that poses many challenges for cost and purchasing managers. On the other hand, ML methods, especially in combination with expert knowledge and other analytical methods, offer immense added value for cost analysis in purchasing.Originality/valueDigital transformation allows to facilitate the cost calculation process in purchasing decisions. In this context, the application of ML approaches has gained increased attention. While such approaches can lead to high cost reductions on the side of both suppliers and buyers, an intelligent cost analysis is very demanding.

Local design and development of low-cost air sensing and data science methods in sub-Saharan Africa – lessons learnt from Uganda

BACKGROUND AND AIM: Sensing and modelling approaches need to be conceived and designed while taking into account unique local contexts in order to be responsive and effective in addressing the environmental data and evidence gaps for the communities and authorities that need them. We report on experiences of designing, developing, deploying and maintaining a local air quality monitoring, modelling and analysis system in Uganda. METHODS: The AirQo system includes a citywide air quality monitoring network of monitors developed locally to reduce maintenance frequency in dusty environments, leverage solar power, have capability for near-real time monitoring and transmission of air quality data over cellular networks, support for complementary static and mobile (motorcycle) deployment approaches, and application of machine learning approaches for calibration to improve data accuracy. RESULTS:The air quality datasets have enabled design and development of data science products that generate evidence to inform awareness, education and policy interventions including: (1) a digital air quality platform targeted at the city governments in Uganda to visualise trends, forecast and spatial variations of air pollution to the local level. (2) public-facing digital channels such as mobile apps and web tools that provide realtime and historical analysis of air pollution in places of interest, for examples, schools. We also report on experiences from community engagement initiatives where citizens can report observed pollution events to the AirQo platform using social media tools such as WhatsApp. CONCLUSIONS:The lessons from this talk are expected to inform ongoing efforts in other African cities and build local capacity in air quality monitoring monitoring and analysis. KEYWORDS: air quality, monitoring, modelling and analysis

Philympics 2021: Prophage Predictions Perplex Programs

Background Most bacterial genomes contain integrated bacteriophages—prophages—in various states of decay. Many are active and able to excise from the genome and replicate, while others are cryptic prophages, remnants of their former selves. Over the last two decades, many computational tools have been developed to identify the prophage components of bacterial genomes, and it is a particularly active area for the application of machine learning approaches. However, progress is hindered and comparisons thwarted because there are no manually curated bacterial genomes that can be used to test new prophage prediction algorithms. Methods We present a library of gold-standard bacterial genome annotations that include manually curated prophage annotations, and a computational framework to compare the predictions from different algorithms. We use this suite to compare all extant stand-alone prophage prediction algorithms to identify their strengths and weaknesses. We provide a FAIR dataset for prophage identification, and demonstrate the accuracy, precision, recall, and f1 score from the analysis of seven different algorithms for the prediction of prophages. Results We identified different strengths and weaknesses between the prophage prediction tools. Several tools exhibit exceptional f1 scores, while others have better recall at the expense of more false positives. The tools vary greatly in runtime performance with few exhibiting all desirable qualities for large-scale analyses. Conclusions Our library of gold-standard prophage annotations and benchmarking framework provide a valuable resource for exploring strengths and weaknesses of current and future prophage annotation tools. We discuss caveats and concerns in this analysis, how those concerns may be mitigated, and avenues for future improvements. This framework will help developers identify opportunities for improvement and test updates. It will also help users in determining the tools that are best suited for their analysis.

Multi defect detection and analysis of electron microscopy images with deep learning

Electron microscopy is widely used to explore defects in crystal structures, but human detecting of defects is often time-consuming, error-prone, and unreliable, and is not scalable to large numbers of images or real-time analysis. In this work, we discuss the application of machine learning approaches to find the location and geometry of different defect clusters in irradiated steels. We show that a deep learning based Faster R-CNN analysis system has a performance comparable to human analysis with relatively small training data sets. This study proves the promising ability to apply deep learning to assist the development of automated microscopy data analysis even when multiple features are present and paves the way for fast, scalable, and reliable analysis systems for massive amounts of modern electron microscopy data.

Application of clustering and the Hungarian algorithm to the problem of consistent vortex tracking in incompressible flowfields

The ability to track vortices spatially and temporally is of great interest for the study of complex and turbulent flows. A methodology to solve the problem of vortex tracking by the application of machine learning approaches is investigated. First a well-known vortex detection algorithm is applied to identify coherent structures. Hierarchical clustering is then conducted followed by a unique application of the Hungarian assignment algorithm. Application to a synthetic flowfield of merging Batchelor vortices results in robust vortex labelling even in a vortex merging event. A robotic PIV experimental dataset of a canonical Ahmed body is used to demonstrate the applicability of the method to three-dimensional flows.Graphic abstract

First-principles prediction of electronic transport in fabricated semiconductor heterostructures via physics-aware machine learning

First-principles techniques for electronic transport property prediction have seen rapid progress in recent years. However, it remains a challenge to predict properties of heterostructures incorporating fabrication-dependent variability. Machine-learning (ML) approaches are increasingly being used to accelerate design and discovery of new materials with targeted properties, and extend the applicability of first-principles techniques to larger systems. However, few studies exploited ML techniques to characterize relationships between local atomic structures and global electronic transport coefficients. In this work, we propose an electronic-transport-informatics (ETI) framework that trains on ab initio models of small systems and predicts thermopower of fabricated silicon/germanium heterostructures, matching measured data. We demonstrate application of ML approaches to extract important physics that determines electronic transport in semiconductor heterostructures, and bridge the gap between ab initio accessible models and fabricated systems. We anticipate that ETI framework would have broad applicability to diverse materials classes.

Application of machine learning approaches to administrative claims data to predict clinical outcomes in medical and surgical patient populations.

ObjectiveThis study aimed to develop and validate a claims-based, machine learning algorithm to predict clinical outcomes across both medical and surgical patient populations.MethodsThis retrospective, observational cohort study, used a random 5% sample of 770,777 fee-for-service Medicare beneficiaries with an inpatient hospitalization between 2009–2011. The machine learning algorithms tested included: support vector machine, random forest, multilayer perceptron, extreme gradient boosted tree, and logistic regression. The extreme gradient boosted tree algorithm outperformed the alternatives and was the machine learning method used for the final risk model. Primary outcome was 30-day mortality. Secondary outcomes were: rehospitalization, and any of 23 adverse clinical events occurring within 30 days of the index admission date.ResultsThe machine learning algorithm performance was evaluated by both the area under the receiver operating curve (AUROC) and Brier Score. The risk model demonstrated high performance for prediction of: 30-day mortality (AUROC = 0.88; Brier Score = 0.06), and 17 of the 23 adverse events (AUROC range: 0.80–0.86; Brier Score range: 0.01–0.05). The risk model demonstrated moderate performance for prediction of: rehospitalization within 30 days (AUROC = 0.73; Brier Score: = 0.07) and six of the 23 adverse events (AUROC range: 0.74–0.79; Brier Score range: 0.01–0.02). The machine learning risk model performed comparably on a second, independent validation dataset, confirming that the risk model was not overfit.Conclusions and relevanceWe have developed and validated a robust, claims-based, machine learning risk model that is applicable to both medical and surgical patient populations and demonstrates comparable predictive accuracy to existing risk models.

Feature Transformation for Corporate Tax Default Prediction: Application of Machine Learning Approaches

Applications of machine learning (ML) and data science have extended significantly into contemporary accounting and finance. Yet, the prediction and analysis of taxpayers’ status are relatively untapped to date. Moreover, this paper focuses on the combination of feature transformation as a novel domain of research for corporate firms’ tax status prediction with the applicability of ML approaches. The paper also applies a tax payment dataset of Finish limited liability firms with failed and non-failed tax information. Seven different ML approaches train across four datasets, transformed to non-transformed, that effectively discriminate the non-default tax firms from their default counterparts. The findings advocate tax administration to choose the single best ML approach and feature transformation method for the execution purpose.

Machine learning for material characterization with an application for predicting mechanical properties

AbstractCurrently, the growth of material data from experiments and simulations is expanding beyond processable amounts. This makes the development of new data‐driven methods for the discovery of patterns among multiple lengthscales and time‐scales and structure‐property relationships essential. These data‐driven approaches show enormous promise within materials science. The following review covers machine learning (ML) applications for metallic material characterization. Many parameters associated with the processing and the structure of materials affect the properties and the performance of manufactured components. Thus, this study is an attempt to investigate the usefulness of ML methods for material property prediction. Material characteristics such as strength, toughness, hardness, brittleness, or ductility are relevant to categorize a material or component according to their quality. In industry, material tests like tensile tests, compression tests, or creep tests are often time consuming and expensive to perform. Therefore, the application of ML approaches is considered helpful for an easier generation of material property information. This study also gives an application of ML methods on small punch test (SPT) data for the determination of the property ultimate tensile strength for various materials. A strong correlation between SPT data and tensile test data was found which ultimately allows to replace more costly tests by simple and fast tests in combination with ML.

Importance of Systematic Right Ventricular Assessment in Cardiac Resynchronization Therapy Candidates: A Machine Learning Approach

Despite all having systolic heart failure and broad QRS intervals, patients screened for cardiac resynchronization therapy (CRT) are highly heterogeneous, and it remains extremely challenging to predict the impact of CRT devices on left ventricular function and outcomes. The aim of this study was to evaluate the relative impact of clinical, electrocardiographic, and echocardiographic data on the left ventricular remodeling and prognosis of CRT candidates by the application of machine learning approaches. One hundred ninety-three patients with systolic heart failure receiving CRT according to current recommendations were prospectively included in this multicenter study. A combination of the Boruta algorithm and random forest methods was used to identify features predicting both CRT volumetric response and prognosis. Model performance was tested using the area under the receiver operating characteristic curve. The k-medoid method was also applied to identify clusters of phenotypically similar patients. From 28 clinical, electrocardiographic, and echocardiographic variables, 16 features were predictive of CRT response, and 11 features were predictive of prognosis. Among the predictors of CRT response, eight variables (50%) pertained to right ventricular size or function. Tricuspid annular plane systolic excursion was the main feature associated with prognosis. The selected features were associated with particularly good prediction of both CRT response (area under the curve, 0.81; 95% CI, 0.74-0.87) and outcomes (area under the curve, 0.84; 95% CI, 0.75-0.93). An unsupervised machine learning approach allowed the identification of two phenogroups of patients who differed significantly in clinical variables and parameters of biventricular size and right ventricular function. The two phenogroups had significantly different prognosis (hazard ratio, 4.70; 95% CI, 2.1-10.0; P<.0001; log-rank P<.0001). Machine learning can reliably identify clinical and echocardiographic features associated with CRT response and prognosis. The evaluation of both right ventricular size and functional parameters has pivotal importance for the risk stratification of CRT candidates and should be systematically performed in patients undergoing CRT.

Machine learning applied on the district heating and cooling sector: a review

Driven by the continuous growing demand for heating and cooling, district heating and cooling systems (DHC) play a major role in the field of energy by providing environmentally friendly solutions for citizens with significant economic impact. Taken also into account the global need for greener and smarter cities, optimization and automation of current DHC operation is more imminent than ever. In order to achieve a transformation of DHC systems, new data-driven technologies are being adopted to reach the goals. In this paper the findings of a systematic literature review are presented covering articles published in the last decades in which the authors described the development and application of machine learning approaches to the DHC sector. In total, 74 articles were retrieved, analysed and categorized into two main categories: (i) heating load/demand prediction and (ii) design, maintenance and scheduling. The survey findings are presented and listed in terms of the machine learning techniques mentioned therein (supervised learning, unsupervised learning and reinforcement learning), the specific application domain (load forecast, design, maintenance and scheduling) of each article providing also insights regarding the source data used and the quality of the results.

Application of machine learning approach on improving quality of semi-trailer truck air suspensions

Application of machine learning approach on improving quality of semi-trailer truck air suspensions