关键词： air pollution   particulate matter   machine learning   exposure modeling  

摘要： Estimating air pollution exposure has long been a challenge for environmental health researchers. Technological advances and novel machine learning methods have allowed us to increase the geographic range and accuracy of exposure models, making them a valuable tool in conducting health studies and identifying hotspots of pollution. Here, we have created a prediction model for daily PM2.5 levels in the Greater London area from 1st January 2005 to 31st December 2013 using an ensemble machine learning approach incorporating satellite aerosol optical depth (AOD), land use, and meteorological data. The predictions were made on a 1 km x 1 km scale over 3960 grid cells. The ensemble included predictions from three different machine learners: a random forest (RF), a gradient boosting machine (GBM), and a k-nearest neighbor (KNN) approach. Our ensemble model performed very well, with a ten-fold cross-validated R-2 of 0.828. Of the three machine learners, the random forest outperformed the GBM and KNN. Our model was particularly adept at predicting day-to-day changes in PM2.5 levels with an out-of-sample temporal R-2 of 0.882. However, its ability to predict spatial variability was weaker, with a R-2 of 0.396. We believe this to be due to the smaller spatial variation in pollutant levels in this area.

关键词： Breast cancer   machine learning   computer-aided diagnosis   histopathology  

摘要： Histopathology is a method used for breast cancer diagnosis. Machine learning (ML) methods have achieved success for supervised learning tasks in the medical domain. In this article, we investigate the impact of ML for the diagnosis of breast cancer using histopathology images of conventional photomicroscopy. Cancer diagnosis is the identification of images as cancer or noncancer, and this involves image preprocessing, feature extraction, classification, and performance analysis. In this article, different approaches to perform these necessary steps are reviewed. We find that most ML research for breast cancer diagnosis has been focused on deep learning. Based on inferences from the recent research activities, we discuss how ML methods can benefit conventional microscopy-based breast cancer diagnosis. Finally, we discuss the research gaps of ML approaches for the implementation in a real pathology environment and propose future research guidelines.

关键词： Critical limb ischemia   Healthcare costs   Healthcare resource utilization   Machine learning   Vascular disease   Predictive preventive personalized medicine  

摘要： Background Critical limb ischemia (CLI) is a severe stage of peripheral arterial disease and has a substantial disease and economic burden not only to patients and families, but also to the society and healthcare systems. We aim to develop a personalized prediction model that utilizes baseline patient characteristics prior to CLI diagnosis to predict subsequent 1-year all-cause hospitalizations and total annual healthcare cost, using a novel Bayesian machine learning platform, Reverse Engineering Forward Simulation (TM) (REFS (TM)), to support a paradigm shift from reactive healthcare to Predictive Preventive and Personalized Medicine (PPPM)-driven healthcare. Methods Patients >= 50 years with CLI plus clinical activity for a 6-month pre-index and a 12-month post-index period or death during the post-index period were included in this retrospective cohort of the linked Optum-Humedica databases. REFS (TM) built an ensemble of 256 predictive models to identify predictors of all-cause hospitalizations and total annual all-cause healthcare costs during the 12-month post-index interval. Results The mean age of 3189 eligible patients was 71.9 years. The most common CLI-related comorbidities were hypertension (79.5%), dyslipidemia (61.4%), coronary atherosclerosis and other heart disease (42.3%), and type 2 diabetes (39.2%). Post-index CLI-related healthcare utilization included inpatient services (14.6%) and >= 1 outpatient visits (32.1%). Median annual all-cause and CLI-related costs per patient were $30,514 and $2196, respectively. REFS (TM) identified diagnosis of skin and subcutaneous tissue infections, cellulitis and abscess, use of nonselective beta-blockers, other aftercare, and osteoarthritis as high confidence predictors of all-cause hospitalizations. The leading predictors for total all-cause costs included region of residence and comorbid health conditions including other diseases of kidney and ureters, blindness of vision defects, chronic ulcer of skin, and ch

关键词： Wheat seed vigor   Near-infrared spectroscopy   Machine learning   Multivariate data analysis   Multiple classification  

摘要： Knowledge of the seed vigor status of individual wheat kernels could provide scientific evidence for the screening of excellent germplasm and the breeding of seedlings. Although many factors collaborate to reduce or render seed vigor, many methods have been employed to detect individual kernel vigor. This study aims to demonstrate the feasibility for using near-infrared (NIR) spectroscopy to detect individual wheat seed vigor and determine suitable machine learning classification models. For this study, 1152 wheat kernel samples were selected, and five-sixths of the portion was treated by artificial aging (AA). All seeds spectra were acquired using a single-seed near-infrared system covering the spectral range of 1200-2400 nm. After NIR spectral collection, all kernels underwent a germination test to confirm their vigor. The spectral data from kernels within 3 germination days, 5 germination days and the non-germination kernels were further used for the development of three-category classification models. After pretreatment by using Savitzky-Golay (SG) second derivative-method and standard normal variate (SNV) correction, the high-dimension spectral data were smoothed, and then were reduced to select most effective wavelengths by two spectral dimensional reduction algorithms: principal component analysis (PCA) and successive projections algorithm (SPA). Four machine learning methodologies, support vector machine (SVM), extreme learning machine (ELM), random forest (RF) and adaptive boosting (AdaBoost) were combined with the two spectral dimensional reduction algorithms to build eight models to discriminate and predict each wheat kernel's vigor. The results demonstrated that the eight three-category machine learning classification models developed with the two spectral dimensional reduction algorithms provided comparable results for individual wheat kernel vigor. The accuracies of the eight models were higher than 84.0%, and PCA-ELM and SPA-RF models afforded the two

关键词： Marbling score   Carcass traits   Machine learning   GWAS   SMOTE   Genetic algorithm   Hanwoo Beef Cattle  

摘要： Pricing of Hanwoo beef in the Korean market is primarily based on meat quality, and particularly on marbling score. The ability to accurately predict marbling score early in the life of an animal is extremely valuable for producers to meet the requirements of their target market, and for genetic selection. A total of 3989 Korean Hanwoo cattle (2108 with 50 k SNP genotypes) and 45 phenotypic features were available for this study. Four machine learning (ML) algorithms were applied to predict six carcass traits and compared against linear regression prediction models. In most scenarios, SMO was the best performing algorithm. The most and least accurately predicted traits were carcass weight and marbling score with correlation of 0.95 and 0.64 respectively. Additionally, the value of using a synthetic minority over-sampling technique (SMOTE) was evaluated and results showed a slight improvement in the prediction error of marbling score. Machine Learning approaches can be useful tools to predict important carcass traits in beef cattle.

关键词： Biotite   Mineral formula   Principal component regression   Oxidation state   Ti-oxy substitution  

摘要： We present a new machine learning method for calculating biotite (sensu lato) structural formula from electron microprobe analysis (EMPA) data, which is based on principal components regression (PCR) of a dataset consisting of 155 fully analyzed biotite references that have chemistry and crystal structure refinement. The dataset is randomly grouped into a training set (75% in amount) and a test set (25% in amount). The training set is used to implement the structural formula and the test set is used to evaluate the performance of the model. The resulting linear regression coefficient matrix is then applied to calculate mole proportions of cations and anions of biotite samples using their compositional data from EMPA. Through this method, the distribution of the different cations and anions in the different sites can be calculated, including the tetrahedral Fe3+, octahedral Fe2+, octahedral Fe3+, OH and O-w(2-) at the O(4) site. The O(4) site is assumed to be occupied by anions with a relation of 2 = F + Cl + OH + O-w(2-). Octahedral and interlayer vacancies could also be estimated in this model. The prediction quality for major elements is perfect with R-2 > 0.95. The absolute errors in the estimated octahedral Fe2+, octahedral Al and OH at O(4) site are determined to be +/- 0.2 apfu (atom per formula unit based on 110 + 2(F, Cl, OH, O)), while those in total Fe3+ and O-w(2-) at O(4) site are approximately +/- 0.3 apfu. A funnel-shaped relationship between absolute error in Fe3+/Sigma Fe ratio and FeOT wt% is observed, with the majority falling in the range of +/- 20%. Compared to previous normalization schemes, our model shows significant improvements in estimating Fe3+/Sigma Fe and O-w(2-) at O(4) site. Our model is capable for calculating mineral formulae of common igneous and hydrothermal biotites, but not suitable for those that have been modified in a post-formation oxidation or reduction process. A supplementary Excel spreadsheet is provided that can be easil

关键词： flood   probabilistic forecasting   support vector regression   fuzzy inference   k-nearest neighbors  

摘要： Probabilistic flood forecasting, which provides uncertain information in the forecasting of floods, is practical and informative for implementing flood-mitigation countermeasures. This study adopted various machine learning methods, including support vector regression (SVR), a fuzzy inference model (FIM), and the k-nearest neighbors (k-NN) method, to establish a probabilistic forecasting model. The probabilistic forecasting method is a combination of a deterministic forecast produced using SVR and a probability distribution of forecast errors determined by the FIM and k-NN method. This study proposed an FIM with a modified defuzzification scheme to transform the FIM's output into a probability distribution, and k-NN was employed to refine the probability distribution. The probabilistic forecasting model was applied to forecast flash floods with lead times of 1-3 hours in Yilan River, Taiwan. Validation results revealed the deterministic forecasting to be accurate, and the probabilistic forecasting was promising in view of a forecasted hydrograph and quantitative assessment concerning the confidence level.

关键词： brain imaging   machine learning   network analysis   schizophrenia  

摘要： Objective Schizophrenia is a chronic and debilitating neuropsychiatric disorder. It has been suggested that impaired brain connectivity underlies the pathophysiology of schizophrenia. Network analysis has thus recently emerged in the field of schizophrenia research. Methods We investigated 48 schizophrenia patients and 24 healthy controls using network analysis and a machine learning method. A number of global and nodal network properties were estimated from graphs that were reconstructed using probabilistic brain tractography. These network properties were then compared between groups and used for machine learning to classify schizophrenia patients and healthy controls. Results In classifying schizophrenia patients and healthy controls via network properties, the support vector machine, random forest, naive Bayes, and gradient boosting machine learning models showed an encouraging level of performance. The overall connectivity was revealed as the most significant contributing feature to this classification among the global network properties. Among the nodal network properties, although the relative importance of each region of interest was not identical, there were still some patterns. Conclusion In conclusion, the possibility exists to classify schizophrenia patients and healthy controls using network properties, and we have found that there is a provisional pattern of involved brain regions among patients with schizophrenia.

关键词： Resampling approach   Random subsampling   Bootstrapping   Flood susceptibility   Machine learning  

摘要： Flood susceptibility projections relying on standalone models, with one-time train-test data splitting for model calibration, yields biased results. This study proposed novel integrative flood susceptibility prediction models based on multi-time resampling approaches, random subsampling (RS) and bootstrapping (BT) algorithms, integrated with machine learning models: generalized additive model (GAM), boosted regression tree (BTR) and multivariate adaptive regression splines (MARS). RS and BT algorithms provided 10 runs of data resampling for learning and validation of the models. Then the mean of 10 runs of predictions is used to produce the flood susceptibility maps (BM). This methodology was applied to Ardabil Province on coastal margins of the Caspian Sea which faced destructive floods. The area under curve (AUC) of receiver operating characteristic (ROC) and true skill statistic (TSS) and correlation coefficient (COR) were utilized to evaluate the predictive accuracy of the proposed models. Results demonstrated that resampling algorithms improved the performance of Standalone GAM, MARS and BRT models. Results also revealed that Standalone models had better performance with the BT algorithm compared to the RS algorithm. BT-GAM model attained superior performance in terms of statistical measures (AUC = 0.98, TSS = 0.93, COR = 0.91), followed by BT-MARS (AUC = 0.97, TSS = 0.93, COR = 0.91) and BT-BRT model (AUC = 0.95, TSS = 0.79, COR = 0.79). Results demonstrated that the proposed models outperformed the benchmark models such as Standalone GAM, MARS, BRT, multilayer perceptron (MLP) and support vector machine (SVM). Given the admirable performance of the proposed models in a large scale area, the promising results can be expected from these models for other regions. (C) 2019 Elsevier B.V. All rights reserved.

关键词： Automated isotope identification   Neural networks   Gamma-ray spectroscopy  

摘要： Pattern recognition algorithms such as artificial neural networks (NNs) and convolution neural networks (CNNs) are prime candidates to perform automated gamma-ray spectroscopy. The way these models train and operate mimic how trained spectroscopists identify spectra. These models have shown promise in identifying gamma-ray spectra with large calibration drift and unknown background radiation fields. In this work, two algorithms for mixtures of radioisotopes based on NN and CNN are presented and evaluated.