Boosted Trees Research Articles

Intensity Analysis for Urban Land Use/Land Cover Dynamics Characterization of Ouagadougou and Bobo-Dioulasso in Burkina Faso

Ouagadougou and Bobo-Dioulasso remain the two major urban centers in Burkina Faso with an increasing trend in human footprint. The research aimed at analyzing the Land Use/Land Cover (LULC) dynamics in the two cities between 2003 and 2021 using intensity analysis, which decomposes LULC changes into interval, category and transition levels. The satellite data used for this research were composed of surface reflectance imagery from Landsat 5, Landsat 7 and Landsat 8 acquired from the Google Earth Engine Data Catalogue. The Random Forest, Support Vector Machine and Gradient Tree Boost algorithms were employed to run supervised image classifications for four selected years including 2003, 2009, 2015 and 2021. The results showed that the landscape is changing in both cities due to rapid urbanization. Ouagadougou experienced more rapid changes than Bobo-Dioulasso, with a maximum annual change intensity of 3.61% recorded between 2015 and 2021 against 2.22% in Bobo-Dioulasso for the period 2009–2015. The transition of change was mainly towards built-up areas, which gain targeted bare and agricultural lands in both cities. This situation has led to a 78.12% increase of built-up surfaces in Ouagadougou, while 42.24% of agricultural land area was lost. However, in Bobo-Dioulasso, the built class has increased far more by 140.67%, and the agricultural land areas experienced a gain of 1.38% compared with the 2003 baseline. The study demonstrates that the human footprint is increasing in both cities making the inhabitants vulnerable to environmental threats such as flooding and the effect of an Urban Heat Island, which is information that could serve as guide for sustainable urban land use planning.

Application of AI-Based Models for Flood Water Level Forecasting and Flood Risk Classification

Owing to global climate change, the frequency of disasters has increased twelve-fold, with a corresponding approximately seventeen-fold increase in economic damages over the past six decades. Notably, severe flood damage has been occurring in Asia along the Pacific coast due to extreme weather events, including torrential rains and typhoons, which have been becoming increasingly frequent and prolonging the rainy season. In the Eastern Visayas region, the management and monitoring facilities for flood observation data, as well as the forecasting and warning systems suitable for the local area, are insufficient. The warning system introduced through overseas grants is limited in operation in some areas of the city. Furthermore, although an organization has jurisdiction over flood forecasting and warning, the system's operation is not systematic and is limited. Additionally, there is a shortage of technical manpower. In this study, we utilized deep learning models to forecast flood water levels in the CarayCaray Basin on Biliran Island, located in Eastern Visayas, the Philippines. Additionally, a flood risk classification was applied to evaluate the degree of risk associated with the predicted water levels. The predicted water levels for each model were compared with the observed water level data. The evaluation of the predictive performance of each model resulted in an NRMSE value of 9.48. Moreover, the accuracy of the DNN model was found to be the best among the flood water level prediction models. To implement the flood risk classification, we utilized extreme gradient boosting, random forest, and decision tree models. The application of these models resulted in an F1-score of 0.92 for the extreme gradient boost model, which exhibited the highest accuracy. With an increasing need for disaster (flood) management, AI-based predictive models are anticipated to reduce the damage caused by natural disasters and enhance disaster mitigation systems. Real-time collection of rainfall and water level data enables continuous learning. Furthermore, if a clear flood warning based on learned flood level patterns is issued, preemptive measures can be taken before intense flood damage occurs.

Machine learning can guide suitability of consultation and patient referral through telemedicine for hepatobiliary diseases.

Telemedicine is an evolving tool to provide health-care services. We evaluated the suitability of telemedicine to deliver effective consultation for hepatobiliary disorders. In this prospective study spanning over a year, we interviewed hepatologists delivering the teleconsultations through a pre-validated questionnaire. A consult was deemed suitable based on the physician's judgment in the absence of unplanned hospitalization. We evaluated factors determining the suitability through inferential statistics and machine learning models, namely, extreme gradient boosting (XGB) and decision tree (DT). Of 1118 consultations, 917 (82.0%) were deemed suitable. On univariable analysis, patients with skilled occupation, higher education, out-of-pocket expenses, and diseases such as chronic hepatitis B, C, and non-alcoholic fatty liver disease (NAFLD) without cirrhosis were associated with suitability (P<0.05). Patients with cirrhosis (compensated or decompensated), acute-on-chronic liver failure (ACLF), and biliary obstruction were likely unsuitable (P<0.05). XGB and DT models predicted suitability with an area under the receiver operating curve of 0.808 and 0.780, respectively. DT demonstrated that compensated cirrhosis with higher education or skilled occupation with age <55years had 78% chance of suitability whereas hepatocellular carcinoma, decompensated cirrhosis, and ACLF patients were unsuitable with a 60-95% probability. In non-cirrhotic liver diseases, hepatitis B, C, and NAFLD were suitable, with a probability of 89.7%. Biliary obstruction and previous failure of teleconsultation were unsuitable, with a probability of 70%. Non-cirrhotic portal fibrosis, dyspepsia, and dysphagia not requiring intervention were suitable (probability: 88%). A simple decision tree can guide the referral of unsuitable and the management of suitable patients with hepatobiliary diseases through telemedicine.

Predictive models for the aqueous phase reactivity of inorganic radicals with organic micropollutants

Single-electron transfer (SET) is one of the most common reaction mechanisms for degrading organic micropollutants (OMPs) in advanced oxidation processes. We collected 300 SET reactions (CO3•–, SO4•–, Cl2•–, and Br2•–-mediated) and calculated three key parameters for understanding the SET mechanism: aqueous phase free energies of activation (ΔG‡), free energies of reactions (ΔG), and orbital energy gaps of reactants (EOMPsHOMO-ERadiLUMO). Then, we classified the OMPs according to their structure, developed and evaluated linear energy relationships of the second-order rate constants (k) with ΔG‡, ΔG, or EOMPsHOMO-ERadiLUMO in each class. Considering that a single descriptor cannot capture all the chemical diversity, we combined ΔG‡, ΔG, and EOMPsHOMO-ERadiLUMO as inputs to develop multiple linear regression (MLR) models. Chemical classification is critical to the linear model described above. However, OMPs usually have multiple functional groups, making the classification challenging and uncertain. Therefore, we tried machine learning algorithms to predict k values without chemical classification. We found that decision trees (R2 = 0.88–0.95) and random forest (R2 = 0.90–0.94) algorithms show better performance on the prediction of the k values, whereas boosted tree algorithm cannot make an accurate prediction (R2 = 0.19–0.36). Overall, our study provides a powerful tool to predict the aqueous phase reactivity of OMP to certain radicals without the need for chemical classification.

On the role of politeness in online human–human tutoring

AbstractResearchers have demonstrated that dialogue‐based intelligent tutoring systems (ITS) can be effective in assisting students in learning. However, little research has attempted to explore the necessity of equipping dialogue‐based ITS with one of the most important capabilities of human tutors, that is, maintaining polite interactions with students, which is essential to provide students with a pleasant learning experience. In this study, we examined the role of politeness by analysing a large‐scale real‐world dataset consisting of over 14K online human–human tutorial dialogues. Specifically, we employed linguistic theories of politeness to characterise the politeness levels of tutor–student‐generated utterances, investigated the correlation between the politeness levels of tutors' utterances and students' problem‐solving performance and quantified the power of politeness in predicting students' problem‐solving performance by applying Gradient Tree Boosting. The study results showed that: (i) in the effective tutorial sessions (ie, sessions in which students successfully solved problems), tutors tended to be very polite at the start of a tutorial session and become more direct to guide students as the session progressed; (ii) students with better performance in solving problems tended to be more polite at the beginning and the end of a tutorial session than their counterparts who failed to solve problems; (iii) the correlation between tutors' polite expressions and students' performance was not evident in non‐instructional communication; and (iv) politeness alone cannot adequately reveal students' problem‐solving performance, and thus other factors (eg, sentiment contained in utterances) should also be taken into account. Practitioner notesWhat is already known about this topic Human–human tutoring is acknowledged as an effective instructional method. Polite expression can help strengthen the relationship between tutors and students. Polite expression can promote students' learning achievements in many educational contexts. What this paper adds By considering the students' prior progress on a problem‐based learning task, we demonstrated the extent to which tutors and students display politeness in tutoring dialogues. Tutors' polite expressions might not correlate with students' problem‐solving performance in online human–human tutoring dialogues. Politeness alone was insufficient to predict the students' performance. Implications for practice Tutors might consider using words with positive sentiment values to express politeness to students with prior progress, which might encourage those students to make a further effort. The polite strategy of expressing indirect requests could help tutors mitigate the sense of directness, but this strategy should be carefully used in delivering instructional hints, especially for students without prior progress. To better assist students without prior progress, tutors might consider using more direct expression to explicitly guide students.

Comparison of Machine Learning Algorithms for Somatosensory Functional Mapping (S30.010)

<h3>Objective:</h3> Using free-running intracranial EEG (icEEG) data, we compare the performance of machine learning algorithms with a binary electrocortical stimulation (ECS) outcome to map functional somatosensory (SM) and nonfunctional areas over the frontoparietal regions. <h3>Background:</h3> Artificial intelligence (AI) is the “science and engineering of making intelligent machines.” Despite advances in AI and its transformative role in medicine, it is lagging in the field of epilepsy surgery. IcEEG data are an ideal medium for AI applications. <h3>Design/Methods:</h3> The study included consecutive cases with intractable epilepsy that underwent icEEG monitoring in a single center between 2013–2018, in which standard functional evaluation of the SM cortex was performed. We evaluated the performance of support vector machines (SVMs), simple multilayer perceptron (MLP), bootstrap forests (BF), boosted trees (BT), decision trees (DT), and logistic regression (LR) to identify the SM cortex using previously validated characteristics from 6–10 minutes of resting icEEG data, applying standard common parameters, 10-fold cross-validation, and no expert supervision. <h3>Results:</h3> Seven subjects and 376 contacts were identified. The median threshold for the SM mapping was 2.5 mA and a range of 1.5 to 7 mA. Function-to-other ratio: 1:2.35. SM cortex showed higher EEG band power than non-SM cortex (medians normalized power 0.22 [IQR 0.11–0.93] vs 0.12 [IQR 0.08–0.24] P &lt; 0.0002). The mean (maximum) AUC for MLP, SVM, BF, LR, BT and DT in classifying function versus the other was 87.4% (95.2%), 84.6% (89%), 84.1% (90.6%), 82% (93.8%), 81.7% (94.2%) and 80.13% (90.7%) respectively. <h3>Conclusions:</h3> When analyzing binary SM functional classification without expert input, the performance of AI algorithms is comparable, with MLP then SVM consistently outperforming the others. Our findings suggest that future research and applications should emphasize clinical features and indications rather than algorithms. To translate our findings into clinical practice, experts should investigate new features, refine parameters within an algorithm, and conduct prospective validations. <b>Disclosure:</b> Dr. Alwan has nothing to disclose. Dennis Spencer, MD has nothing to disclose. Rafeed Alkawadri has nothing to disclose.

QUIC Network Traffic Classification Using Ensemble Machine Learning Techniques

The Quick UDP Internet Connections (QUIC) protocol provides advantages over traditional TCP, but its encryption functionality reduces the visibility for operators into network traffic. Many studies deploy machine learning and deep learning algorithms on QUIC traffic classification. However, standalone machine learning models are subject to overfitting and poor predictability in complex network traffic environments. Deep learning on the other hand requires a huge dataset and intensive parameter fine-tuning. On the contrary, ensemble techniques provide reliability, better prediction, and robustness of the trained model, thereby reducing the chance of overfitting. In this paper, we approach the QUIC network traffic classification problem by utilizing five different ensemble machine learning techniques, namely: Random Forest, Extra Trees, Gradient Boosting Tree, Extreme Gradient Boosting Tree, and Light Gradient Boosting Model. We used the publicly available dataset with five different services such as Google Drive, YouTube, Google Docs, Google Search, and Google Music. The models were trained using a different number of features on different scenarios and evaluated using several performance metrics. The results show that Extreme Gradient Boosting Tree and Light Gradient Boosting Model outperform the other models and achieve one of the highest results among the state-of-the-art models found in the literature with a simpler model and features.

A Comprehensive Study on Crop Recommendation System for Precision Agriculture Using Machine Learning Algorithms

Crop recommendation is among the most crucial components of the field known as precision agriculture. The crop recommendations were formulated after considering a wide range of distinct considerations. The term "precision agriculture" refers to a method of contemporary farming that makes use of information on soil features, soil kinds, and other factors, such as crop yields, and weather conditions to provide farmers with recommendations regarding the types of crops that would be most beneficial to grow on their farms to achieve the highest possible levels of both yield and profit. The crop recommendation system is currently being built with the use of machine learning methods including Random Forest, Gradient Boosting, XG Boost, Light GBM, SVM, and decision tree. The process of identifying the best crop to produce is aided by the information that consists of qualities such as potassium (K), phosphorous (P), and nitrogen (N) as well as temperature, humidity, pH, and rainfall. This application will be of assistance to farmers in enhancing agricultural output, minimizing the deterioration of soil on the cultivated ground, reducing the number of chemicals used in agricultural production, and increasing the productivity of cultivated land with available water resources. Also, this application recommends the top 5 prioritized crops to farmers. XG Boost achieved the best accuracy of 99.500 out of all these classifiers that were utilized in this paper. It also achieved the highest recall score of 99.545455, a precision score of 99.564935, and an F1 score of 99.545669.

An integrative machine learning framework for classifying SEER breast cancer

Breast cancer is the commonest type of cancer in women worldwide and the leading cause of mortality for females. The aim of this research is to classify the alive and death status of breast cancer patients using the Surveillance, Epidemiology, and End Results dataset. Due to its capacity to handle enormous data sets systematically, machine learning and deep learning has been widely employed in biomedical research to answer diverse classification difficulties. Pre-processing the data enables its visualization and analysis for use in making important decisions. This research presents a feasible machine learning-based approach for categorizing SEER breast cancer dataset. Moreover, a two-step feature selection method based on Variance Threshold and Principal Component Analysis was employed to select the features from the SEER breast cancer dataset. After selecting the features, the classification of the breast cancer dataset is carried out using Supervised and Ensemble learning techniques such as Ada Boosting, XG Boosting, Gradient Boosting, Naive Bayes and Decision Tree. Utilizing the train-test split and k-fold cross-validation approaches, the performance of various machine learning algorithms is examined. The accuracy of Decision Tree for both train-test split and cross validation achieved as 98%. In this study, it is observed that the Decision Tree algorithm outperforms other supervised and ensemble learning approaches for the SEER Breast Cancer dataset.

61 Detecting Parkinson’s Disease Using Computer Vision

OBJECTIVES/GOALS: Can we detect Parkinson’s-disease-related motor impairments using computer vision and machine learning? METHODS/STUDY POPULATION: A sample of 29 people with Parkinson’s disease (PD) and 29 non-Parkinson’s disease (non-PD) controls were recruited from the University of Iowa Movement Disorders Clinic. Videos of 3 motor assessment tasks performed using the hands were recorded and hand location information was abstracted using the computer vision program MediaPipe. Measures from the raw data series and FFT were used as features to train a model using boosted trees to classify each video as PD or non-PD. Model performance was evaluated using leave-one-out cross-validation. Additionally, we used recursive feature elimination to reduce model complexity. RESULTS/ANTICIPATED RESULTS: A model using two features identified by recursive feature elimination yielded a model with an overall accuracy of 81% in cross-validation. In our sample, the model had 86.2% sensitivity, 75.9% specificity, and an AUC of 0.839. Additional improvement may be possible with more data processing, especially in the time-domain. DISCUSSION/SIGNIFICANCE: We built a classifier that was able to reliably and accurately discriminate between videos of motor assessments in people with Parkinson’s and people without. This may provide a low cost screening tool in rural areas or primary care clinics with limited access to neurologist expertise.

Predictive Analysis of Hospital Stay after Caesarean Section: A Single-Center Study.

Caesarean section (CS) rate has seen a significant increase in recent years, especially in industrialized countries. There are, in fact, several causes that justify a CS; however, evidence is emerging that non-obstetric factors may contribute to the decision. In reality, CS is not a risk-free procedure. The intra-operative, post-pregnancy risks and risks for children are just a few examples. From a cost point of view, it must be considered that CS requires longer recovery times, and women often stay hospitalized for several days. This study analyzed data from 12,360 women who underwent CS at the "San Giovanni di Dio e Ruggi D'Aragona" University Hospital between 2010 and 2020 by multiple regression algorithms, including multiple linear regression (MLR), Random Forest, Gradient Boosted Tree, XGBoost, and linear regression, classification algorithms and neural network in order to study the variation of the dependent variable (total LOS) as a function of a group of independent variables. We identify the MLR model as the most suitable because it achieves an R-value of 0.845, but the neural network had the best performance (R = 0.944 for the training set). Among the independent variables, Pre-operative LOS, Cardiovascular disease, Respiratory disorders, Hypertension, Diabetes, Haemorrhage, Multiple births, Obesity, Pre-eclampsia, Complicating previous delivery, Urinary and gynaecological disorders, and Complication during surgery were the variables that significantly influence the LOS. Among the classification algorithms, the best is Random Forest, with an accuracy as high as 77%. The simple regression model allowed us to highlight the comorbidities that most influence the total LOS and to show the parameters on which the hospital management must focus for better resource management and cost reduction.

Predicting quality parameters of wastewater treatment plants using artificial intelligence techniques

Estimating wastewater treatment plants’ (WWTPs) influent parameters such as 5-day biological oxygen demand (BOD5) and chemical oxygen demand (COD) is vital for optimizing electricity and energy consumption. Against this backdrop, the existing body of knowledge is bereft of a study employing Artificial Intelligence-based techniques for the prediction of BOD5 and COD. Thus, in this study, Gene expression programming (GEP), multilayer perception neural networks, multi-linear regression, k-nearest neighbors, gradient boosting, and regression trees -based models were trained for predicting BOD5 and COD, using monthly data collected from the inflow of 7 WWTPs over a three-year period in Hong Kong. Based on different statistical parameters, GEP provides more accurate estimations, with R2 values of 0.784 and 0.861 for BOD5 and COD respectively. Furthermore, results of sensitivity analysis undertaken by monte Carlo simulation revealed that both BOD5 and COD were mostly affected by concentrations of total suspended solids, and a 10% increase in the value of TSS resulted in a 7.94% and 7.92% increase in the values of BOD5 and COD, respectively. It is seen that the GEP modeling results complied with the fundamental chemistry of the wastewater quality parameters and can be further applied on other sewage sources such as industrial sewage and leachate. The promising results obtained pave the way for forecasting the operational parameters during sludge processing, leading to an extensive energy savings during the wastewater treatment processes.

Evaluation of IIOT based Pd-MaaS using CNN with Ensemble Subspace Discriminate – for Indian Ship Building in Maritime Industry

Indian shipbuilding has a long history in the maritime industry dating back to the origin of civilization. India's shipbuilding sector is primarily concentrated in its coastal regions. Due to capacity constraints and decreased shipbuilding prices in emerging nations, shipbuilding activities has changed. This has created fresh opportunities for the Indian shipbuilding industry. The prospects for the Indian shipbuilding sector are improved by rising global trade and strong need for modern boats. This study investigates the use of Predictive Maintenance as a Service on the Industrial Internet of Things (IIoT-PdMaaS). Artificial intelligence (AI) in the maritime industry has numerous major benefits, including improved decision-making analysis, automation, security, route planning, and increased efficiency. So, Pd-MaaS using IIOT (Convolution neural network (CNN) with Ensemble Boosted Tree Classifier) framework was developed in this study. This research shows 88.3% accuracy of CNN output for confusion matrix implying a positive connection with our proposed model for Indian ship building industry

Application of Data Mining Techniques to Predict Luminance of Pavement Aggregate

The primary purpose of the analysis presented here is to assess the feasibility of effectively predicting the aggregate luminance coefficient. Current road lighting standards and recommendations are based on assessing the level and distribution of luminance on the road surface. The brightness of a road surface depends on the amount of light falling on it, as well as the reflective properties of the road surface, which in turn depend on its physical condition, type and mineralogical composition. The complexity of the factors on which the value of the luminance coefficient depends it makes that data mining techniques the most appropriate tools for evaluation luminance coefficient phenomenon. This article uses five types of techniques: C&amp;RT, boosted trees, random forest, neural network, and support vector machines. After a preliminary analysis, it was determined that the most effective technique was the boosted tree method. The results of the analysis indicated that the actual value of the luminance coefficient has multiple modal values within a single aggregate stockpile, depending on the mineralogical composition and grain size, and cannot be determined by a single central measure. The present model allowed us to determine the value of the luminance coefficient Qd with a mean error of 4.3 mcd-m−2·lx−1. In addition, it was found that the best aggregate for pavement brightening that allows high visibility during the day Qd and at night RL is a limestone aggregate. In the group of those that have the ability to potentially brighten the pavement were quartzite and granite aggregates.

Digital assessments of soil organic carbon storage using digital maps provided by static and dynamic environmental covariates

AbstractUnderstanding the vertical and lateral distribution of soil organic carbon (SOC) and soil organic carbon density (SOCD) is indispensable for soil use and environmental management because of their vital role in soil quality assessments. Primarily, they are needed in calculating soil organic carbon storage (SOCS). The objective of this research was to provide digital maps of SOC and SOCD variation as well as their uncertainties at multiple standardized depths (H1: 0–5, H2: 5–15, H3: 15–30, H4: 30–60 and H5: 60–100 cm) using a parsimonious model with optimized terrain‐related attributes and satellite‐derived data. SOCS were evaluated at soil subgroup levels. An area of about 808 km2 with varying elevation, plant cover and lithology from the Miandoab region, West Azerbaijan Province, Iran was selected as a case study area. A total of 386 soil samples were collected from 104 profiles comprising various soil genetic horizons. A continuous spline function was then fitted to the target properties in advance of creating a dataset at five standard depth intervals (following the GlobalSoilMap project). These were then grouped into three classes including top (H1), middle (H2, H3 and H4) and bottom (H5) depths to ease interpretation. Static and dynamic covariates (30‐m resolution) were derived from a digital elevation model (DEM) and a suite of Landsat‐8 spectral imageries, respectively. Four candidate models including stepwise multiple linear regression (SMLR), random forest (RF), cubist (CU) and extreme gradient boosting (XGBoost) Tree were tested in this study. Finally, the digital maps at 30‐m resolution of SOC and SOCD and their uncertainties were prepared using the best‐fit model and the bootstrapping method, respectively. Four soil subgroups (Gypsic Haploxerepts, Typic Calcixerepts, Typic Haploxerepts and Xeric Haplocalcids) were identified across the study area. The covariates had variable contributions on the evaluated models. The XGBoost Tree model generally outperformed other models for prediction of SOC and SOCD (R2 = 0.60, on average). Regardless of soil subgroups, the uncertainty analysis showed that the SOCD map had a low prediction interval range value indicating high accuracy. Additionally, the highest SOCS and SOCD was observed at the top followed by middle and bottom depths in the study area. All subgroups exhibited a decreasing trend of SOCD with increasing depth. A similar trend was also observed for SOCS. The highest SOCD (on average) was observed in Gypsic Haploxerepts (4.71 kg C/m2) followed by Typic Calcixerepts (4.46 kg C/m2), Typic Haploxerepts (4.45 kg C/m2) and Xeric Haplocalcids (4.40 kg C/m2). Overall, the SOCS normalized by area within soil order boundaries was greater in Inceptisols than Aridisols across the study area. The findings of this study provide critical information for sustainable management of soil resources in the area for agricultural production and environmental health in the Miandoab region of Iran.

Developing Pedestrian Fatality Prediction Models Using Historical Crash Data: Application of Binary Logistic Regression and Boosted Tree Mechanism

Pedestrian fatality rate plays a key role in examining effectiveness of the road safety. The present study attempts to examine the effect of various categories of accused vehicles and the average 85th percentile speed at accident location on the pedestrian crash fatality. The study also attempts to develop pedestrian crash severity models using the binary logistic regression and boosted trees technique. Historical crash data, along with the video recording technique at accident sites, have been utilized for the present study. From regression equations, it is observed that when the heavy vehicle (HV) hits a pedestrian as compared to two-wheeler (2W), the average chance of death increases 2.44 times. According to the Boosted tree model, the contribution of speed is 60 %, whereas the contribution of category of accused vehicle is 40 % for pedestrian fatality prediction. The study should help in planning better strategies like all red time at intersections or pedestrian foot over bridge at critical locations.

Disentangling environmental effects on picophytoplankton communities in the Eastern Indian Ocean

Photosynthesis by picophytoplankton provides energy for higher organisms and is essential in the food chain and global carbon cycle. In 2020 and 2021, we investigated the spatial distribution and vertical changes of picophytoplankton in the euphotic layer of the Eastern Indian Ocean (EIO) and estimated their carbon biomass contributions through two cruise surveys. The abundance of picophytoplankton was composed of Prochlorococcus (69.94%), Synechococcus (22.21%) and picoeukaryotes (7.85%). Synechococcus was mainly found in the surface layer, while Prochlorococcus and picoeukaryotes had high abundances in the subsurface layer. The surface picophytoplankton community was greatly affected by fluorescence, the middle layer was significantly regulated by temperature and dissolved oxygen concentration, and the lower layer was dominated by nutrients and apparent oxygen utilization (AOU). Aggregated boosted tree (ABT) and Generalized Additive models (GAM) indicated that temperature, salinity, AOU, and fluorescence were strong influencing factors of picophytoplankton communities in EIO. The mean carbon biomass contribution of picophytoplankton in the surveyed area was 0.565 μg C/L, which was contributed by Prochlorococcus (39.32%), Synechococcus (38.88%) and picoeukaryotes (21.80%). These findings contribute to our understanding of the effects of different environmental factors on picophytoplankton communities and the influence of picophytoplankton contributions to the carbon pools of the oligotrophic ocean.

High-Throughput FPGA-Based Inference of Gradient Tree Boosting Models for Position Estimation in PET Detectors

In modern large-scale positron emission tomography (PET) systems, transferring the digitized raw detector data at high count rates to a centralized processing unit is a challenge. Processing data on field programmable gate arrays (FPGAs) close to the detectors can reduce data early on and improve scalability of the PET system. We present and evaluate an FPGA implementation of gradient tree boosting (GTB) for 1-D position estimation of gamma interactions in the scintillator. GTB is a supervised machine learning algorithm based on building ensembles of binary decision trees. Models were trained offline and inferred in an FPGA (XC7K410T-2FFG676 Kintex-7). Input features and GTB parameters influencing both positioning performance and model size were varied while evaluating the inferred models concerning data throughput and FPGA resource consumption as well as positioning performance. We achieved throughputs per detector between <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$2.94\times 10^{6}$ </tex-math></inline-formula> and <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$4.55\times 10^{6}$ </tex-math></inline-formula> gamma interactions per second. For an optimized GTB model, resource consumption could be reduced by factors of 17 and 10 to less than 1% ( <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$2.51\times 10^{3}$ </tex-math></inline-formula> look-up tables) of available logic and 1.26% (20 BRAMs) of memory resources, while maintaining a positioning performance of 98.63% when compared to the model with the best positioning performance. The presented framework can be easily adapted to other photosensors and scintillator influencing.

Full-length 16S rRNA gene sequencing and machine learning reveal the bacterial composition of inhalable particles from two different breeding stages in a piggery

Bacterial loading aggravates the harm of particulate matter (PM) to public health and ecological systems, especially in operations of concentrated animal production. This study aimed to explore the characteristics and influencing factors of bacterial components of inhalable particles at a piggery. The morphology and elemental composition of coarse particles (PM10, aerodynamic diameter ≤ 10 µm) and fine particles (PM2.5, aerodynamic diameter ≤ 2.5 µm) were analyzed. Full-length 16 S rRNA sequencing technology was used to identify bacterial components according to breeding stage, particle size, and diurnal rhythm. Machine learning (ML) algorithms were used to further explore the relationship between bacteria and the environment. The results showed that the morphology of particles in the piggery differed, and the morphologies of the suspected bacterial components were elliptical deposited particles. Full-length 16 S rRNA indicated that most of the airborne bacteria in the fattening and gestation houses were bacilli. The analysis of beta diversity and difference between samples showed that the relative abundance of some bacteria in PM2.5 was significantly higher than that in PM10 at the same pig house (P < 0.01). There were significant differences in the bacterial composition of inhalable particles between the fattening and gestation houses (P < 0.01). The aggregated boosted tree (ABT) model showed that PM2.5 had a great influence on airborne bacteria among air pollutants. Fast expectation-maximization microbial source tracking (FEAST) showed that feces was a major potential source of airborne bacteria in pig houses (contribution 52.64–80.58 %). These results will provide a scientific basis for exploring the potential risks of airborne bacteria in a piggery to human and animal health.

Comprehensible Machine-Learning-Based Models for the Pre-Emptive Diagnosis of Multiple Sclerosis Using Clinical Data: A Retrospective Study in the Eastern Province of Saudi Arabia

Multiple Sclerosis (MS) is characterized by chronic deterioration of the nervous system, mainly the brain and the spinal cord. An individual with MS develops the condition when the immune system begins attacking nerve fibers and the myelin sheathing that covers them, affecting the communication between the brain and the rest of the body and eventually causing permanent damage to the nerve. Patients with MS (pwMS) might experience different symptoms depending on which nerve was damaged and how much damage it has sustained. Currently, there is no cure for MS; however, there are clinical guidelines that help control the disease and its accompanying symptoms. Additionally, no specific laboratory biomarker can precisely identify the presence of MS, leaving specialists with a differential diagnosis that relies on ruling out other possible diseases with similar symptoms. Since the emergence of Machine Learning (ML) in the healthcare industry, it has become an effective tool for uncovering hidden patterns that aid in diagnosing several ailments. Several studies have been conducted to diagnose MS using ML and Deep Learning (DL) models trained using MRI images, achieving promising results. However, complex and expensive diagnostic tools are needed to collect and examine imaging data. Thus, the intention of this study is to implement a cost-effective, clinical data-driven model that is capable of diagnosing pwMS. The dataset was obtained from King Fahad Specialty Hospital (KFSH) in Dammam, Saudi Arabia. Several ML algorithms were compared, namely Support Vector Machine (SVM), Decision Tree (DT), Logistic Regression (LR), Random Forest (RF), Extreme Gradient Boosting (XGBoost), Adaptive Boosting (AdaBoost), and Extra Trees (ET). The results indicated that the ET model outpaced the rest with an accuracy of 94.74%, recall of 97.26%, and precision of 94.67%.