scholarly journals Applying Machine Learning and AI Explanations to Analyze Vaccine Hesitancy

Author(s):  
Carsten Lange ◽  
Jian Lange

The paper identifies and quantifies the impact of race, poverty, politics, and age on COVID-19 vaccination rates in counties across the continental US. Both traditional Ordinary Least Square (OLS) regression analysis and Random Forest machine learning algorithms are applied to quantify contributing factors for county-level vaccination hesitancy. With the machine learning model, joint effects of multiple variables (race/ethnicity, partisanship, age etc.) are considered simultaneously to capture the unique combination of what factors affect the vaccination rate. By implementing a state-of-the-art Artificial Intelligence Explanations (AIX) algorithm, it is possible to solve the black box problem with machine learning models and provide answers to the "how much" question for each measured impact factor in every county. For most counties a higher percentage vote for Republicans, a greater African American population share, and a higher poverty rate lower the vaccination rate. While a higher Asian population share increases the predicted vaccination rate. The impact on the vaccination rate from the Hispanic population proportion is positive in the OLS model, but only positive for counties with very high Hispanic population (65% and more) in the Random Forest model. Both the proportion of seniors and the one for young people in a county have a significant impact in the OLS model - positive and negative, respectively. In contrast, the impacts are ambiguous in the Random Forest model. Because results vary between geographies and since the AIX algorithm is able to quantify vaccine impacts individually for each county, this research can be tailored to local communities. This way it is a helpful tool for local health officials and other policymakers to improve vaccination rates. An interactive online mapping dashboard that identifies impact factors for individual U.S. counties is available at https://www.cpp.edu/~clange/vacmap.html. It is apparent that the influence of impact factors is not universally the same across different geographies.

2019 ◽  
Author(s):  
Manesh Chawla ◽  
Amreek Singh

Abstract. Fast downslope release of snow (avalanche) is a serious hazard to people living in snow bound mountains. Released snow mass can gain sufficient momentum on its down slope path to kill humans, uproot trees and rocks, destroy buildings. Direct reduction of avalanche threat is done by building control structures to add mechanical support to snowpack and reduce or deflect downward avalanche flow. On large terrains it is economically infeasible to use these methods on each high risk site.Therefore predicting and avoiding avalanches is the only feasible method to reduce threat but sufficient snow stability data for accurate forecasting is generally unavailable and difficult to collect. Forecasters infer snow stability from their knowledge of local weather, terrain and sparsely available snowpack observations. This inference process is vulnerable to human bias therefore machine learning models are used to find patterns from past data and generate helpful outputs to minimise and quantify uncertainty in forecasting process. These machine learning techniques require long past records of avalanches which are difficult to obtain. In this paper we propose a data efficient Random Forest model to address this problem. The model can generate a descriptive forecast showing reasoning and patterns which are difficult to observe manually. Our model advances the field by being inexpensive and convenient for operational forecasting due to its data efficiency, ease of automation and ability to describe its decisions.


Cancers ◽  
2021 ◽  
Vol 13 (23) ◽  
pp. 6013
Author(s):  
Hyun-Soo Park ◽  
Kwang-sig Lee ◽  
Bo-Kyoung Seo ◽  
Eun-Sil Kim ◽  
Kyu-Ran Cho ◽  
...  

This prospective study enrolled 147 women with invasive breast cancer who underwent low-dose breast CT (80 kVp, 25 mAs, 1.01–1.38 mSv) before treatment. From each tumor, we extracted eight perfusion parameters using the maximum slope algorithm and 36 texture parameters using the filtered histogram technique. Relationships between CT parameters and histological factors were analyzed using five machine learning algorithms. Performance was compared using the area under the receiver-operating characteristic curve (AUC) with the DeLong test. The AUCs of the machine learning models increased when using both features instead of the perfusion or texture features alone. The random forest model that integrated texture and perfusion features was the best model for prediction (AUC = 0.76). In the integrated random forest model, the AUCs for predicting human epidermal growth factor receptor 2 positivity, estrogen receptor positivity, progesterone receptor positivity, ki67 positivity, high tumor grade, and molecular subtype were 0.86, 0.76, 0.69, 0.65, 0.75, and 0.79, respectively. Entropy of pre- and postcontrast images and perfusion, time to peak, and peak enhancement intensity of hot spots are the five most important CT parameters for prediction. In conclusion, machine learning using texture and perfusion characteristics of breast cancer with low-dose CT has potential value for predicting prognostic factors and risk stratification in breast cancer patients.


Author(s):  
Qian Zhao ◽  
Ning Xu ◽  
Hui Guo ◽  
Jianguo Li

Background: Sepsis is a life-threatening disease caused by the dysregulated host response to the infection, and being the major cause of death to patients in intensive care unit (ICU). Objective: Early diagnosis of sepsis could significantly reduce in-hospital mortality. Though generated from infection, the development of sepsis follows its own psychological process and disciplines, alters with gender, health status and other factors. Hence, the analysis of mass data by bioinformatic tools and machine learning is a promising method for exploring early diagnosis manners. Method: We collected miRNA and mRNA expression data of sepsis blood samples from Gene Expression Omnibus (GEO) and ArrayExpress databases, screened out differentially expressed genes (DEGs) by R software, predicted miRNA targets on TargetScanHuman and miRTarBase websites, conducted Gene Ontology (GO) term and KEGG pathway enrichment based on overlapping DEGs. The STRING database and Cytoscape were used to build protein-protein interaction (PPI) network and predict hub genes. Then we constructed a Random Forest model by using the hub genes to assess sample type. Results: Bioinformatic analysis of GEO dataset revealed 46 overlapping DEGs in sepsis. The PPI network analysis identified five hub genes, SOCS3, KBTBD6, FBXL5, FEM1C and WSB1. Random Forest model based on these five hub genes was used to assess GSE95233 and GSE95233 datasets, and the area under curve (AUC) of ROC are 0.900 and 0.7988, respectively, which confirmed the efficacy of this model. Conclusion: The integrated analysis of gene expression in sepsis and the effective Random Forest model built in this study may provide promising diagnostic methods for sepsis.


2019 ◽  
Vol 9 (1) ◽  
Author(s):  
Eun Kyung Park ◽  
Kwang-sig Lee ◽  
Bo Kyoung Seo ◽  
Kyu Ran Cho ◽  
Ok Hee Woo ◽  
...  

AbstractRadiogenomics investigates the relationship between imaging phenotypes and genetic expression. Breast cancer is a heterogeneous disease that manifests complex genetic changes and various prognosis and treatment response. We investigate the value of machine learning approaches to radiogenomics using low-dose perfusion computed tomography (CT) to predict prognostic biomarkers and molecular subtypes of invasive breast cancer. This prospective study enrolled a total of 723 cases involving 241 patients with invasive breast cancer. The 18 CT parameters of cancers were analyzed using 5 machine learning models to predict lymph node status, tumor grade, tumor size, hormone receptors, HER2, Ki67, and the molecular subtypes. The random forest model was the best model in terms of accuracy and the area under the receiver-operating characteristic curve (AUC). On average, the random forest model had 13% higher accuracy and 0.17 higher AUC than the logistic regression. The most important CT parameters in the random forest model for prediction were peak enhancement intensity (Hounsfield units), time to peak (seconds), blood volume permeability (mL/100 g), and perfusion of tumor (mL/min per 100 mL). Machine learning approaches to radiogenomics using low-dose perfusion breast CT is a useful noninvasive tool for predicting prognostic biomarkers and molecular subtypes of invasive breast cancer.


Viruses ◽  
2020 ◽  
Vol 12 (2) ◽  
pp. 142 ◽  
Author(s):  
Steven J. Erly ◽  
Joshua T. Herbeck ◽  
Roxanne P. Kerani ◽  
Jennifer R. Reuer

Molecular cluster detection can be used to interrupt HIV transmission but is dependent on identifying clusters where transmission is likely. We characterized molecular cluster detection in Washington State, evaluated the current cluster investigation criteria, and developed a criterion using machine learning. The population living with HIV (PLWH) in Washington State, those with an analyzable genotype sequences, and those in clusters were described across demographic characteristics from 2015 to2018. The relationship between 3- and 12-month cluster growth and demographic, clinical, and temporal predictors were described, and a random forest model was fit using data from 2016 to 2017. The ability of this model to identify clusters with future transmission was compared to Centers for Disease Control and Prevention (CDC) and the Washington state criteria in 2018. The population with a genotype was similar to all PLWH, but people in a cluster were disproportionately white, male, and men who have sex with men. The clusters selected for investigation by the random forest model grew on average 2.3 cases (95% CI 1.1–1.4) in 3 months, which was not significantly larger than the CDC criteria (2.0 cases, 95% CI 0.5–3.4). Disparities in the cases analyzed suggest that molecular cluster detection may not benefit all populations. Jurisdictions should use auxiliary data sources for prediction or continue using established investigation criteria.


2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Silvia Alonso ◽  
Sara Cáceres ◽  
Daniel Vélez ◽  
Luis Sanz ◽  
Gema Silvan ◽  
...  

AbstractSteroidal hormone interaction in pregnancy is crucial for adequate fetal evolution and preparation for childbirth and extrauterine life. Estrone sulphate, estriol, progesterone and cortisol play important roles in the initiation of labour mechanism at the start of contractions and cervical effacement. However, their interaction remains uncertain. Although several studies regarding the hormonal mechanism of labour have been reported, the prediction of date of birth remains a challenge. In this study, we present for the first time machine learning algorithms for the prediction of whether spontaneous labour will occur from week 37 onwards. Estrone sulphate, estriol, progesterone and cortisol were analysed in saliva samples collected from 106 pregnant women since week 34 by enzyme-immunoassay (EIA) techniques. We compared a random forest model with a traditional logistic regression over a dataset constructed with the values observed of these measures. We observed that the results, evaluated in terms of accuracy and area under the curve (AUC) metrics, are sensibly better in the random forest model. For this reason, we consider that machine learning methods contribute in an important way to the obstetric practice.


2021 ◽  
Vol 6 (1) ◽  
pp. 295-309
Author(s):  
Daniel Vassallo ◽  
Raghavendra Krishnamurthy ◽  
Harindra J. S. Fernando

Abstract. Machine learning is quickly becoming a commonly used technique for wind speed and power forecasting. Many machine learning methods utilize exogenous variables as input features, but there remains the question of which atmospheric variables are most beneficial for forecasting, especially in handling non-linearities that lead to forecasting error. This question is addressed via creation of a hybrid model that utilizes an autoregressive integrated moving-average (ARIMA) model to make an initial wind speed forecast followed by a random forest model that attempts to predict the ARIMA forecasting error using knowledge of exogenous atmospheric variables. Variables conveying information about atmospheric stability and turbulence as well as inertial forcing are found to be useful in dealing with non-linear error prediction. Streamwise wind speed, time of day, turbulence intensity, turbulent heat flux, vertical velocity, and wind direction are found to be particularly useful when used in unison for hourly and 3 h timescales. The prediction accuracy of the developed ARIMA–random forest hybrid model is compared to that of the persistence and bias-corrected ARIMA models. The ARIMA–random forest model is shown to improve upon the latter commonly employed modeling methods, reducing hourly forecasting error by up to 5 % below that of the bias-corrected ARIMA model and achieving an R2 value of 0.84 with true wind speed.


2021 ◽  
Author(s):  
Hemalatha N ◽  
Akhil Wilson ◽  
Akhil Thankachan

Plastic pollution is one of the challenging problems in the environment. But a life without plastic we cannot imagine. This paper deals with the prediction of plastic degrading microbes using Machine Learning. Here we have used Decision Tree, Random Forest, Support vector Machine and K Nearest Neighbor algorithms in order to predict the plastic degrading microbes. Among the four classifiers, Random Forest model gave the best accuracy of 99.1%.


2021 ◽  
Vol 14 ◽  
pp. 175628482110531
Author(s):  
Asaf Levartovsky ◽  
Yiftach Barash ◽  
Shomron Ben-Horin ◽  
Bella Ungar ◽  
Shelly Soffer ◽  
...  

Background: Intra-abdominal abscess (IA) is an important clinical complication of Crohn’s disease (CD). A high index of clinical suspicion is needed as imaging is not routinely used during hospital admission. This study aimed to identify clinical predictors of an IA among hospitalized patients with CD using machine learning. Methods: We created an electronic data repository of all patients with CD who visited the emergency department of our tertiary medical center between 2012 and 2018. We searched for the presence of an IA on abdominal imaging within 7 days from visit. Machine learning models were trained to predict the presence of an IA. A logistic regression model was compared with a random forest model. Results: Overall, 309 patients with CD were hospitalized and underwent abdominal imaging within 7 days. Forty patients (12.9%) were diagnosed with an IA. On multivariate analysis, high C-reactive protein (CRP) [above 65 mg/l, adjusted odds ratio (aOR): 16 (95% CI: 5.51–46.18)], leukocytosis [above 10.5 K/μl, aOR: 4.47 (95% CI: 1.91–10.45)], thrombocytosis [above 322.5 K/μl, aOR: 4.1 (95% CI: 2–8.73)], and tachycardia [over 97 beats per minute, aOR: 2.7 (95% CI: 1.37–5.3)] were independently associated with an IA. Random forest model showed an area under the curve of 0.817 ± 0.065 with six features (CRP, hemoglobin, WBC, age, current biologic therapy, and BUN). Conclusion: In our large tertiary center cohort, the machine learning model identified the association of six clinical features (CRP, hemoglobin, WBC, age, BUN, and biologic therapy) with the presentation of an IA. These may assist as a decision support tool in triaging CD patients for imaging to exclude this potentially life-threatening complication.


Sign in / Sign up

Export Citation Format

Share Document