DroughtCast: A Machine Learning Forecast of the United States Drought Monitor.
Ontology highlight
ABSTRACT:
SUBMITTER: Brust C
PROVIDER: S-EPMC8725730 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:BackgroundMeasles, a highly contagious viral infection, is resurging in the United States, driven by international importation and declining domestic vaccination coverage. Despite this resurgence, measles outbreaks are still rare events that are difficult to predict. Improved methods to predict outbreaks at the county level would facilitate the optimal allocation of public health resources.ObjectiveWe aimed to validate and compare extreme gradient boosting (XGBoost) and logistic regression, 2 supervised learning approaches, to predict the US counties most likely to experience measles cases. We also aimed to assess the performance of hybrid versions of these models that incorporated additional predictors generated by 2 clustering algorithms, hierarchical density-based spatial clustering of applications with noise (HDBSCAN) and unsupervised random forest (uRF).MethodsWe constructed a supervised machine learning model based on XGBoost and unsupervised models based on HDBSCAN and uRF. The unsupervised models were used to investigate clustering patterns among counties with measles outbreaks; these clustering data were also incorporated into hybrid XGBoost models as additional input variables. The machine learning models were then compared to logistic regression models with and without input from the unsupervised models.ResultsBoth HDBSCAN and uRF identified clusters that included a high percentage of counties with measles outbreaks. XGBoost and XGBoost hybrid models outperformed logistic regression and logistic regression hybrid models, with the area under the receiver operating curve values of 0.920-0.926 versus 0.900-0.908, the area under the precision-recall curve values of 0.522-0.532 versus 0.485-0.513, and F2 scores of 0.595-0.601 versus 0.385-0.426. Logistic regression or logistic regression hybrid models had higher sensitivity than XGBoost or XGBoost hybrid models (0.837-0.857 vs 0.704-0.735) but a lower positive predictive value (0.122-0.141 vs 0.340-0.367) and specificity (0.793-0.821 vs 0.952-0.958). The hybrid versions of the logistic regression and XGBoost models had slightly higher areas under the precision-recall curve, specificity, and positive predictive values than the respective models that did not include any unsupervised features.ConclusionsXGBoost provided more accurate predictions of measles cases at the county level compared with logistic regression. The threshold of prediction in this model can be adjusted to align with each county's resources, priorities, and risk for measles. While clustering pattern data from unsupervised machine learning approaches improved some aspects of model performance in this imbalanced data set, the optimal approach for the integration of such approaches with supervised machine learning models requires further investigation.
| S-EPMC10131820 | biostudies-literature
Project description:We explore the feasibility of using machine learning on a police dataset to forecast domestic homicides. Existing forecasting instruments based on ordinary statistical instruments focus on non-fatal revictimization, produce outputs with limited predictive validity, or both. We implement a "super learner," a machine learning paradigm that incorporates roughly a dozen machine learning models to increase the recall and AUC of forecasting using any one model. We purposely incorporate police records only, rather than multiple data sources, to illustrate the practice utility of the super learner, as additional datasets are often unavailable due to confidentiality considerations. Using London Metropolitan Police Service data, our model outperforms all extant domestic homicide forecasting tools: the super learner detects 77.64% of homicides, with a precision score of 18.61% and a 71.04% Area Under the Curve (AUC), which, collectively and severely, are assessed as "excellent." Implications for theory, research, and practice are discussed.
| S-EPMC10739734 | biostudies-literature
Project description:The massive socioeconomic impacts engendered by extreme floods provides a clear motivation for improved understanding of flood drivers. We use self-organizing maps, a type of artificial neural network, to perform unsupervised clustering of climate reanalysis data to identify synoptic-scale atmospheric circulation patterns associated with extreme floods across the United States. We subsequently assess the flood characteristics (e.g., frequency, spatial domain, event size, and seasonality) specific to each circulation pattern. To supplement this analysis, we have developed an interactive website with detailed information for every flood of record. We identify four primary categories of circulation patterns: tropical moisture exports, tropical cyclones, atmospheric lows or troughs, and melting snow. We find that large flood events are generally caused by tropical moisture exports (tropical cyclones) in the western and central (eastern) United States. We identify regions where extreme floods regularly occur outside the normal flood season (e.g., the Sierra Nevada Mountains due to tropical moisture exports) and regions where multiple extreme flood events can occur within a single year (e.g., the Atlantic seaboard due to tropical cyclones and atmospheric lows or troughs). These results provide the first machine-learning based near-continental scale identification of atmospheric circulation patterns associated with extreme floods with valuable insights for flood risk management.
| S-EPMC6509142 | biostudies-literature
Project description:Since 2013, the Centers for Disease Control and Prevention (CDC) has hosted an annual influenza season forecasting challenge. The 2015-2016 challenge consisted of weekly probabilistic forecasts of multiple targets, including fourteen models submitted by eleven teams. Forecast skill was evaluated using a modified logarithmic score. We averaged submitted forecasts into a mean ensemble model and compared them against predictions based on historical trends. Forecast skill was highest for seasonal peak intensity and short-term forecasts, while forecast skill for timing of season onset and peak week was generally low. Higher forecast skill was associated with team participation in previous influenza forecasting challenges and utilization of ensemble forecasting techniques. The mean ensemble consistently performed well and outperformed historical trend predictions. CDC and contributing teams will continue to advance influenza forecasting and work to improve the accuracy and reliability of forecasts to facilitate increased incorporation into public health response efforts.
| S-EPMC6346105 | biostudies-literature
Project description:PurposeHepatocellular carcinoma (HCC) incidence rates have been increasing in the United States for the past 35 years. Because HCC has a poor prognosis, quantitative forecasts could help to inform prevention and treatment strategies to reduce the incidence and burden of HCC.MethodsSingle-year HCC incident case and population data for the years 2000 to 2012 and ages 35 to 84 years were obtained from the SEER 18 Registry Database. We forecast incident HCC cases through 2030, using novel age-period-cohort models and stratifying by sex, race/ethnicity, and age. Rates are presented because absolute numbers may be influenced by population increases.ResultsRates of HCC increased with each successive birth cohort through 1959. However, rates began to decrease with the 1960 to 1969 birth cohorts. Asians/Pacific Islanders (APIs) have had the highest HCC rates in the United States for many years, but the rates have stabilized and begun to decline in recent years. Between 2013 and 2030, rates among APIs are forecast to decline further, with estimated annual percentage changes of -1.59% among men and -2.20% among women. Thus, by 2030, Asians are forecast to have the lowest incidence rates among men, and Hispanics are forecast to have the highest rates among men (age-standardized rate, 44.2). Blacks are forecast to have the highest rate among women (age-standardized rate, 12.82).ConclusionAlthough liver cancer has long had some of the most rapidly increasing incidence rates, the decreasing rates seen among APIs, individuals younger than 65 years, and cohorts born after 1960 suggest that there will be declines in incidence of HCC in future years. Prevention efforts should be focused on individuals in the 1950 to 1959 birth cohorts, Hispanics, and blacks.
| S-EPMC4966339 | biostudies-literature
Project description:Background: This study aimed to better characterize morbidly obese kidney transplant recipients, their clinical characteristics, and outcomes by using an unsupervised machine learning approach. Methods: Consensus cluster analysis was applied to OPTN/UNOS data from 2010 to 2019 based on recipient, donor, and transplant characteristics in kidney transplant recipients with a pre-transplant BMI ≥ 40 kg/m2. Key cluster characteristics were identified using the standardized mean difference. Post-transplant outcomes, including death-censored graft failure, patient death, and acute allograft rejection, were compared among the clusters. Results: Consensus clustering analysis identified 3204 kidney transplant recipients with a BMI ≥ 40 kg/m2. In this cohort, five clinically distinct clusters were identified. Cluster 1 recipients were predominantly white and non-sensitized, had a short dialysis time or were preemptive, and were more likely to receive living donor kidney transplants. Cluster 2 recipients were older and diabetic. They were likely to have been on dialysis >3 years and receive a standard KDPI deceased donor kidney. Cluster 3 recipients were young, black, and had kidney disease secondary to hypertension or glomerular disease. Cluster 3 recipients had >3 years of dialysis and received non-ECD, young, deceased donor kidney transplants with a KDPI < 85%. Cluster 4 recipients were diabetic with variable dialysis duration who either received non-ECD standard KDPI kidneys or living donor kidney transplants. Cluster 5 recipients were young retransplants that were sensitized. One-year patient survival in clusters 1, 2, 3, 4, and 5 was 98.0%, 94.4%, 98.5%, 98.7%, and 97%, and one-year death-censored graft survival was 98.1%, 93.0%, 96.1%, 98.8%, and 93.0%, respectively. Cluster 2 had the worst one-year patient survival. Clusters 2 and 5 had the worst one-year death-censored graft survival. Conclusions: With the application of unsupervised machine learning, variable post-transplant outcomes are observed among morbidly obese kidney transplant recipients. Recipients with earlier access to transplant and living donation show superior outcomes. Unexpectedly, reduced graft survival in cluster 3 recipients perhaps underscores socioeconomic access to post-transplant support and minorities being disadvantaged in access to preemptive and living donor transplants. Despite obesity-related concerns, one-year patient and graft survival were favorable in all clusters, and obesity itself should be reconsidered as a hard barrier to kidney transplantation.
| S-EPMC9224965 | biostudies-literature
Project description:Recent studies have shown that systems combining mathematical modeling and Bayesian inference methods can be used to generate real-time forecasts of future infectious disease incidence. Here we develop such a system to study and forecast respiratory syncytial virus (RSV). RSV is the most common cause of acute lower respiratory infection and bronchiolitis. Advanced warning of the epidemic timing and volume of RSV patient surges has the potential to reduce well-documented delays of treatment in emergency departments. We use a susceptible-infectious-recovered (SIR) model in conjunction with an ensemble adjustment Kalman filter (EAKF) and ten years of regional U.S. specimen data provided by the Centers for Disease Control and Prevention. The data and EAKF are used to optimize the SIR model and i) estimate critical epidemiological parameters over the course of each outbreak and ii) generate retrospective forecasts. The basic reproductive number, R0, is estimated at 3.0 (standard deviation 0.6) across all seasons and locations. The peak magnitude of RSV outbreaks is forecast with nearly 70% accuracy (i.e. nearly 70% of forecasts within 25% of the actual peak), four weeks before the predicted peak. This work represents a first step in the development of a real-time RSV prediction system.
| S-EPMC5055361 | biostudies-literature
Project description:BackgroundTesticular germ cell tumors (TGCTs) are rare tumors in the general population but are the most commonly occurring malignancy among males between ages 15 and 44 years in the United States (US). Although non-Hispanic whites (NHWs) have the highest incidence in the US, rates among Hispanics have increased the most in recent years. To forecast what these incidence rates may be in the future, an analysis of TGCT incidence in the Surveillance, Epidemiology, and End Results program and the National Program of Cancer Registries was conducted.MethodsTGCT incidence data among males ages 15 to 59 years for the years 1999 to 2012 were obtained from 39 US cancer registries. Incidence rates through 2026 were forecast using age-period-cohort models stratified by race/ethnicity, histology (seminoma, nonseminoma), and age.ResultsBetween 1999 and 2012, TGCT incidence rates, both overall and by histology, were highest among NHWs, followed by Hispanics, Asian/Pacific Islanders, and non-Hispanic blacks. Between 2013 and 2026, rates among Hispanics were forecast to increase annually by 3.96% (95% confidence interval, 3.88%-4.03%), resulting in the highest rate of increase of any racial/ethnic group. By 2026, the highest TGCT rates in the US will be among Hispanics because of increases in both seminomas and nonseminomas. Rates among NHWs will slightly increase, whereas rates among other groups will slightly decrease.ConclusionsBy 2026, Hispanics will have the highest rate of TGCT of any racial/ethnic group in the US because of the rising incidence among recent birth cohorts. Reasons for the increase in younger Hispanics merit further exploration. Cancer 2017;123:2320-2328. © 2017 American Cancer Society.
| S-EPMC5629636 | biostudies-literature
Project description:The COVID-19 pandemic has been widely spread and affected millions of people and caused hundreds of deaths worldwide, especially in patients with comorbilities and COVID-19. This manuscript aims to present models to predict, firstly, the number of coronavirus cases and secondly, the hospital care demand and mortality based on COVID-19 patients who have been diagnosed with other diseases. For the first part, I present a projection of the spread of coronavirus in Mexico, which is based on a contact tracing model using Bayesian inference. I investigate the health profile of individuals diagnosed with coronavirus to predict their type of patient care (inpatient or outpatient) and survival. Specifically, I analyze the comorbidity associated with coronavirus using Machine Learning. I have implemented two classifiers: I use the first classifier to predict the type of care procedure that a person diagnosed with coronavirus presenting chronic diseases will obtain (i.e. outpatient or hospitalised), in this way I estimate the hospital care demand; I use the second classifier to predict the survival or mortality of the patient (i.e. survived or deceased). I present two techniques to deal with these kinds of unbalanced datasets related to outpatient/hospitalised and survived/deceased cases (which occur in general for these types of coronavirus datasets) to obtain a better performance for the classification.
| S-EPMC8782335 | biostudies-literature
Project description:Water is stored in reservoirs for various purposes, including regular distribution, flood control, hydropower generation, and meeting the environmental demands of downstream habitats and ecosystems. However, these objectives are often in conflict with each other and make the operation of reservoirs a complex task, particularly during flood periods. An accurate forecast of reservoir inflows is required to evaluate water releases from a reservoir seeking to provide safe space for capturing high flows without having to resort to hazardous and damaging releases. This study aims to improve the informed decisions for reservoirs management and water prerelease before a flood occurs by means of a method for forecasting reservoirs inflow. The forecasting method applies 1- and 2-month time-lag patterns with several Machine Learning (ML) algorithms, namely Support Vector Machine (SVM), Artificial Neural Network (ANN), Regression Tree (RT), and Genetic Programming (GP). The proposed method is applied to evaluate the performance of the algorithms in forecasting inflows into the Dez, Karkheh, and Gotvand reservoirs located in Iran during the flood of 2019. Results show that RT, with an average error of 0.43% in forecasting the largest reservoirs inflows in 2019, is superior to the other algorithms, with the Dez and Karkheh reservoir inflows forecasts obtained with the 2-month time-lag pattern, and the Gotvand reservoir inflow forecasts obtained with the 1-month time-lag pattern featuring the best forecasting accuracy. The proposed method exhibits accurate inflow forecasting using SVM and RT. The development of accurate flood-forecasting capability is valuable to reservoir operators and decision-makers who must deal with streamflow forecasts in their quest to reduce flood damages.
| S-EPMC8692612 | biostudies-literature