Advanced Machine Learning Methods Research Articles

Causal Inference in Strategy: How can Machine Learning Help?

Recent advances in machine learning methods have opened up a set of new possibilities for inductive theorizing, and for analysis of unstructured text data in strategy and organizations research. In contrast, the application of machine learning to improve causal explanation in deductive strategy research is less well understood. In this paper, we spotlight supervised machine learning and its application to matching methods to support stronger inference in deductive research. We use a simulation and an analysis of technology invention data to illustrate the method. A core contribution is to provide guidance to strategy researchers in the use of supervised machine learning.

Algorithmic Face-ism: Uncovering and Mitigating Algorithmic Bias in Facial Recognition Systems

How do some of the most advanced machine learning facial recognition algorithms make important decisions, such as whom to hire or who is considered a leader? Existing research suggests advances in machine learning methods can use facial features in an image—facial morphology—to accurately and objectively predict answers to such questions. We show, however, that even after implementing state-of-the-art models, decision-based facial recognition algorithms are not as objective as previously claimed. Unpacking the “black box” of an existing facial recognition algorithm revealed the algorithm did not rely on facial morphology to make decisions. Instead, when covariates such as attractiveness were accounted for, the algorithm relied mostly on “leftover” transient features, such as clothing, hairstyle, or background lighting for decision-making. We identify the specific stages—sampling, preprocessing, model implementation, and model functioning stage— in which algorithmic focus bias and interpretation bias are likely to arise in facial recognition algorithms. These results suggest that decision-based facial recognition algorithms are biased in ways that researchers have overlooked, with troubling implications for their use by governments, organizations, and researchers. We introduce the concept of “algorithmic face-ism,” in which (1) machine learning algorithms unfairly express an inherent preference for specific facial morphologies, and (2) researchers mistakenly attribute behavioral predictions to facial morphologies. This paper thus demonstrates how leading decision-based facial recognition systems are biased and how previously taken-for-granted factors contribute to this pattern of bias. We conclude by discussing how bias can be mitigated in such facial recognition algorithms.

Predicting apneic events in preterm infants using cardio-respiratory and movement features

Background and objectivePreterm neonates are prone to episodes of apnea, bradycardia and hypoxia (ABH) that can lead to neurological morbidities or even death. There is broad interest in developing methods for real-time prediction of ABH events to inform interventions that prevent or reduce their incidence and severity. Using advances in machine learning methods, this study develops an algorithm to predict ABH events. MethodsFollowing previous studies showing that respiratory instabilities are closely associated with bouts of movement, we present a modeling framework that can predict ABH events using both movement and cardio-respiratory features derived from routine clinical recordings. In 10 preterm infants, movement onsets and durations were estimated with a wavelet-based algorithm that quantified artifactual distortions of the photoplethysmogram signal. For prediction, cardio-respiratory features were created from time-delayed correlations of inter-beat and inter-breath intervals with past values; movement features were derived from time-delayed correlations with inter-breath intervals. Gaussian Mixture Models and Logistic Regression were used to develop predictive models of apneic events. Performance of the models was evaluated with ROC curves. ResultsPerformance of the prediction framework (mean AUC) was 0.77 ± 0.04 for 66 ABH events on training data from 7 infants. When grouped by the severity of the associated bradycardia during the ABH event, the framework was able to predict 83% and 75% of the most severe episodes in the 7-infant training set and 3-infant test set, respectively. Notably, inclusion of movement features significantly improved the predictions compared with modeling with only cardio-respiratory signals. ConclusionsOur findings suggest that recordings of movement provide important information for predicting ABH events in preterm infants, and can inform preemptive interventions designed to reduce the incidence and severity of ABH events.

Pandemic Analytics by Advanced Machine Learning for Improved Decision Making of COVID-19 Crisis

With the advent of the first pandemic wave of Severe Acute Respiratory Syndrome Coronavirus-2 (SARS-CoV-2), the question arises as to whether the spread of the virus will be controlled by the application of preventive measures or will follow a different course, regardless of the pattern of spread already recorded. These conditions caused by the unprecedented pandemic have highlighted the importance of reliable data from official sources, their complete recording and analysis, and accurate investigation of epidemiological indicators in almost real time. There is an ongoing research demand for reliable and effective modeling of the disease but also the formulation of substantiated views to make optimal decisions for the design of preventive or repressive measures by those responsible for the implementation of policy in favor of the protection of public health. The main objective of the study is to present an innovative data-analysis system of COVID-19 disease progression in Greece and her border countries by real-time statistics about the epidemiological indicators. This system utilizes visualized data produced by an automated information system developed during the study, which is based on the analysis of large pandemic-related datasets, making extensive use of advanced machine learning methods. Finally, the aim is to support with up-to-date technological means optimal decisions in almost real time as well as the development of medium-term forecast of disease progression, thus assisting the competent bodies in taking appropriate measures for the effective management of the available health resources.

Applications of machine learning methods in traffic crash severity modelling: current status and future directions

ABSTRACT As a key area of traffic safety research, crash severity modelling has attracted tremendous attention. Recently, there has been growing interest in applying machine learning (ML) methods in this area. However, the lessons and experience learned so far have not been systematically documented and summarised. This is the first article that surveys studies on ML applications in crash severity modelling and has the following major contributions: (1) it provides a comprehensive and critical review of current research efforts; (2) it summarises the successful experience and main challenges (e.g. data and methodology); and (3) it identifies promising research opportunities towards accurate and reliable crash severity modelling and results interpretation. The review results suggest that imbalanced data remains a major issue. Under- and over-samplings are often used to balance crash severity data despite their limitations. Some studies use local sensitivity analysis (LSA) to interpret ML modelling results but ignore the strict assumptions of LSA and omit the joint effects of risk factors. Moreover, very few studies consider the accuracy and reliability of ML model evaluation metrics. Other issues include spatiotemporal correlations, causality, model transferability and heterogeneity. This paper concludes by providing suggestions on model selection and modification to address the identified issues and recommendations for future research. For example, employing advanced ML methods such as graph convolutional networks (GCN) to model spatiotemporal correlations; exploring innovative ways of applying ML methods; and leveraging new developments in ML (e.g. interpretable ML) to derive causal relationships and interpret modelling results.

Product sales probabilistic forecasting: An empirical evaluation using the M5 competition data

Supply chain management depends heavily on uncertain point forecasts of product sales. In order to deal with such uncertainty and optimize safety stock levels, methods that can estimate the right part of the sales distribution are required. Given the limited work that has been done in the field of probabilistic product sales forecasting, we propose and test some novel methods to estimate uncertainty, utilizing empirical computations and simulations to determine quantiles. To do so, we use the M5 competition data to empirically evaluate the forecasting and inventory performance of these methods by making comparisons both with established statistical approaches and advanced machine learning methods. Our results indicate that different methods should be employed based on the quantile of interest and the characteristics of the series being forecast, concluding that methods that employ relatively simple and faster to compute empirical estimations result in better inventory performance than more sophisticated and computer intensive approaches.

Prediction of 30-Day Readmission After Stroke Using Machine Learning and Natural Language Processing.

Background and Purpose: This study aims to determine whether machine learning (ML) and natural language processing (NLP) from electronic health records (EHR) improve the prediction of 30-day readmission after stroke.Methods: Among index stroke admissions between 2011 and 2016 at an academic medical center, we abstracted discrete data from the EHR on demographics, risk factors, medications, hospital complications, and discharge destination and unstructured textual data from clinician notes. Readmission was defined as any unplanned hospital admission within 30 days of discharge. We developed models to predict two separate outcomes, as follows: (1) 30-day all-cause readmission and (2) 30-day stroke readmission. We compared the performance of logistic regression with advanced ML algorithms. We used several NLP methods to generate additional features from unstructured textual reports. We evaluated the performance of prediction models using a five-fold validation and tested the best model in a held-out test dataset. Areas under the curve (AUCs) were used to compare discrimination of each model.Results: In a held-out test dataset, advanced ML methods along with NLP features out performed logistic regression for all-cause readmission (AUC, 0.64 vs. 0.58; p < 0.001) and stroke readmission prediction (AUC, 0.62 vs. 0.52; p < 0.001).Conclusion: NLP-enhanced machine learning models potentially advance our ability to predict readmission after stroke. However, further improvement is necessary before being implemented in clinical practice given the weak discrimination.

Are Machine Learning Methods the Future for Smoking Cessation Apps?

Smoking cessation apps provide efficient, low-cost and accessible support to smokers who are trying to quit smoking. This article focuses on how up-to-date machine learning algorithms, combined with the improvement of mobile phone technology, can enhance our understanding of smoking behaviour and support the development of advanced smoking cessation apps. In particular, we focus on the pros and cons of existing approaches that have been used in the design of smoking cessation apps to date, highlighting the need to improve the performance of these apps by minimizing reliance on self-reporting of environmental conditions (e.g., location), craving status and/or smoking events as a method of data collection. Lastly, we propose that making use of more advanced machine learning methods while enabling the processing of information about the user’s circumstances in real time is likely to result in dramatic improvement in our understanding of smoking behaviour, while also increasing the effectiveness and ease-of-use of smoking cessation apps, by enabling the provision of timely, targeted and personalised intervention.

Health and sleep nursing assistant for real-time, contactless, and non-invasive monitoring

In this paper, we introduce a novel health and sleep nursing assistant called “Helena” for real-time, contactless, and non-invasive monitoring that can be mounted on a bed frame to continuously monitor sleep activities (entry/exit of bed, movement, and posture changes), vital signs (heart rate and respiration rate), and falls from bed in a pervasive computing manner. The smart sensor senses bed vibrations generated by body movements to characterize sleep activities and vital signs based on advanced signal processing and machine learning methods. The device can provide information about sleep patterns, generate real-time results, and support continuous sleep assessment and health tracking. The novel method for detecting falls from bed has not been attempted before and represents a life-changing for high-risk communities, such as seniors. Comprehensive tests and validations were conducted to evaluate system performances using FDA approved and wearable devices. Our system has an accuracy of 99.5% detecting on-bed (entries), 99.73% detecting off-bed (exits), 97.92% detecting movements on the bed, 92.08% detecting posture changes, and 97% detecting falls from bed. The system estimation of heart rate (HR) ranged ±2.41 beats-per-minute compared to Apple Watch Series 4, while the respiration rate (RR) ranged ±0.89 respiration-per-minute compared to an FDA (Food and Drug Administration) oximeter and a metronome.

Improving Risk Identification of Adverse Outcomes in Chronic Heart Failure Using SMOTE+ENN and Machine Learning.

PurposeThis study sought to develop models with good identification for adverse outcomes in patients with heart failure (HF) and find strong factors that affect prognosis.Patients and MethodsA total of 5004 qualifying cases were selected, among which 498 cases had adverse outcomes and 4506 cases were discharged after improvement. The study subjects were hospitalized patients diagnosed with HF from a regional cardiovascular hospital and the cardiology department of a medical university hospital in Shanxi Province of China between January 2014 and June 2019. Synthesizing minority oversampling technology combined with edited nearest neighbors (SMOTE+ENN) was used to pre-process unbalanced data. Traditional logistic regression (LR), k-nearest neighbor (KNN), support vector machine (SVM), random forest (RF), and extreme gradient boosting (XGBoost) were used to build risk identification models, and each model was repeated 100 times. Model discrimination and calibration were estimated using F1-score, the area under the receiver-operating characteristic curve (AUROC), and Brier score. The best performing of the five models was used to identify the risk of adverse outcomes and evaluate the influencing factors.ResultsThe SME-XGBoost was the best performing model with means of F1-score (0.3673, 95% confidence interval [CI]: 0.3633–0.3712), AUC (0.8010, CI: 0.7974–0.8046), and Brier score (0.1769, CI: 0.1748–0.1789). Age, N-terminal pronatriuretic peptide, pulmonary disease, etc. were the most significant factors of adverse outcomes in patients with HF.ConclusionThe combination of SMOTE+ENN and advanced machine learning methods effectively improved the discrimination efficacy of adverse outcomes in HF patients, accurately stratified patients at risk of adverse outcomes, and found the top factors of adverse outcomes. These models and factors emphasize the importance of health status data in determining adverse outcomes in patients with HF.

Machine learning-based risk profile classification of patients undergoing elective heart valve surgery.

Machine learning methods potentially provide a highly accurate and detailed assessment of expected individual patient risk before elective cardiac surgery. Correct anticipation of this risk allows for the improved counselling of patients and avoidance of possible complications. We therefore investigated the benefit of modern machine learning methods in personalized risk prediction for patients undergoing elective heart valve surgery. We performed a monocentric retrospective study in patients who underwent elective heart valve surgery between 1 January 2008 and 31 December 2014 at our centre. We used random forests, artificial neural networks and support vector machines to predict the 30-day mortality from a subset of 129 available demographic and preoperative parameters. Exclusion criteria were reoperation of the same patient, patients who needed anterograde cerebral perfusion due to aortic arch surgery and patients with grown-up congenital heart disease. Finally, the cohort consisted of 2229 patients with a 30-day mortality of 3.86% (86 of 2229 cases). This trial has been registered at clinicaltrials.gov (NCT03724123). The final random forest model trained on the entire data set provided an out-of-bag area under the receiver operator characteristics curve (AUC) of 0.839, which significantly outperformed the European System for Cardiac Operative Risk Evaluation (EuroSCORE) (AUC = 0.704) and a model trained only on the subset of features EuroSCORE uses (AUC = 0.745). Advanced machine learning methods can predict outcomes of valve surgery procedures with higher accuracy than established risk scores based on logistic regression on pre-selected parameters. This approach is generalizable to other elective high-risk interventions and allows for training models to the cohorts of specific institutions.

Prediction of Major Complications and Readmission After Lumbar Spinal Fusion: A Machine Learning–Driven Approach

Given the significant cost and morbidity of patients undergoing lumbar fusion, accurate preoperative risk-stratification would be of great utility. We aim to develop a machine learning model for prediction of major complications and readmission after lumbar fusion. We also aim to identify the factors most important to performance of each tested model. We identified 38,788 adult patients who underwent lumbar fusion at any California hospital between 2015 and 2017. The primary outcome was major perioperative complication or readmission within 30 days. We build logistic regression and advanced machine learning models: XGBoost, AdaBoost, Gradient Boosting, and Random Forest. Discrimination and calibration were assessed using area under the receiver operating characteristic curve and Brier score, respectively. There were 4470 major complications (11.5%). The XGBoost algorithm demonstrates the highest discrimination of the machine learning models, outperforming regression. The variables most important to XGBoost performance include angina pectoris, metastatic cancer, teaching hospital status, history of concussion, comorbidity burden, and workers' compensation insurance. Teaching hospital status and concussion history were not found to be important for regression. We report a machine learning algorithm for prediction of major complications and readmission after lumbar fusion that outperforms logistic regression. Notably, the predictors most important for XGBoost differed from those for regression. The superior performance of XGBoost may be due to the ability of advanced machine learning methods to capture relationships between variables that regression is unable to detect. This tool may identify and address potentially modifiable risk factors, helping risk-stratify patients and decrease complication rates.

Predicting youth diabetes risk using NHANES data and machine learning

Prediabetes and diabetes mellitus (preDM/DM) have become alarmingly prevalent among youth in recent years. However, simple questionnaire-based screening tools to reliably assess diabetes risk are only available for adults, not youth. As a first step in developing such a tool, we used a large-scale dataset from the National Health and Nutritional Examination Survey (NHANES) to examine the performance of a published pediatric clinical screening guideline in identifying youth with preDM/DM based on American Diabetes Association diagnostic biomarkers. We assessed the agreement between the clinical guideline and biomarker criteria using established evaluation measures (sensitivity, specificity, positive/negative predictive value, F-measure for the positive/negative preDM/DM classes, and Kappa). We also compared the performance of the guideline to those of machine learning (ML) based preDM/DM classifiers derived from the NHANES dataset. Approximately 29% of the 2858 youth in our study population had preDM/DM based on biomarker criteria. The clinical guideline had a sensitivity of 43.1% and specificity of 67.6%, positive/negative predictive values of 35.2%/74.5%, positive/negative F-measures of 38.8%/70.9%, and Kappa of 0.1 (95%CI: 0.06–0.14). The performance of the guideline varied across demographic subgroups. Some ML-based classifiers performed comparably to or better than the screening guideline, especially in identifying preDM/DM youth (p = 5.23 × 10−5).We demonstrated that a recommended pediatric clinical screening guideline did not perform well in identifying preDM/DM status among youth. Additional work is needed to develop a simple yet accurate screener for youth diabetes risk, potentially by using advanced ML methods and a wider range of clinical and behavioral health data.

ATTICUS: Ambient-Intelligent Tele-monitoring and Telemetry for Incepting and Catering Over hUman Sustainability

Wearable devices as medical technologies are becoming an integral part of our lives. Many research studies are dedicated to these devices and are mainly focused on providing personal analytics, measuring physical status, and acquiring physiological signals and parameters. These continuously evolving technologies play an important role in telemedicine. Telemedicine can be broadly defined as the use of advanced telecommunications technologies to support many medical activities, such as the diagnosis, the analysis of patient data, the improvement of disease management and the treatment in remote areas. In this article, we present ATTICUS (Ambient-intelligent Tele-monitoring and Telemetry for Incepting and Catering over hUman Sustainability), an innovative remote monitoring system for ambient-assisted living based on the analysis of vital and behavioral parameters. The ATTICUS system consists of two essential components: a smart wearable—in the form of a short singlet—made of innovative textile which allows the acquisition of real-time body signals, e.g., electrocardiogram (ECG), breathing wave, temperature, and a multi-level Decision Support System (DSS), a distributed software which integrates advanced machine learning methods to automatically detect anomalies. ATTICUS is capable of operating in different application scenarios. Especially, the system will support in-home and out-home monitoring, personal check-ups, and specialized check-ups. Thus, the system will positively impact the canonical medical practices allowing simultaneous and continuous monitoring of a large number of people.

Variational Disentanglement for Rare Event Modeling

Combining the increasing availability and abundance of healthcare data and the current advances in machine learning methods have created renewed opportunities to improve clinical decision support systems. However, in healthcare risk prediction applications, the proportion of cases with the condition (label) of interest is often very low relative to the available sample size. Though very prevalent in healthcare, such imbalanced classification settings are also common and challenging in many other scenarios. So motivated, we propose a variational disentanglement approach to semi-parametrically learn from rare events in heavily imbalanced classification problems. Specifically, we leverage the imposed extreme-distribution behavior on a latent space to extract information from low-prevalence events, and develop a robust prediction arm that joins the merits of the generalized additive model and isotonic neural nets. Results on synthetic studies and diverse real-world datasets, including mortality prediction on a COVID-19 cohort, demonstrate that the proposed approach outperforms existing alternatives.

Experimental Verification of Self-Adapting Data-Driven Controllers in Active Distribution Grids

Lately, data-driven algorithms have been proposed to design local controls for Distributed Generators (DGs) that can emulate the optimal behaviour without any need for communication or centralised control. The design is based on historical data, advanced off-line optimization techniques and machine learning methods, and has shown great potential when the operating conditions are similar to the training data. However, safety issues arise when the real-time conditions start to drift away from the training set, leading to the need for online self-adapting algorithms and experimental verification of data-driven controllers. In this paper, we propose an online self-adapting algorithm that adjusts the DG controls to tackle local power quality issues. Furthermore, we provide experimental verification of the data-driven controllers through power Hardware-in-the-Loop experiments using an industrial inverter. The results presented for a low-voltage distribution network show that data-driven schemes can emulate the optimal behaviour and the online modification scheme can mitigate local power quality issues.

Pure Ion Chromatograms Combined with Advanced Machine Learning Methods Improve Accuracy of Discriminant Models in LC-MS-Based Untargeted Metabolomics.

Untargeted metabolomics based on liquid chromatography coupled with mass spectrometry (LC–MS) can detect thousands of features in samples and produce highly complex datasets. The accurate extraction of meaningful features and the building of discriminant models are two crucial steps in the data analysis pipeline of untargeted metabolomics. In this study, pure ion chromatograms were extracted from a liquor dataset and left-sided colon cancer (LCC) dataset by K-means-clustering-based Pure Ion Chromatogram extraction method version 2.0 (KPIC2). Then, the nonlinear low-dimensional embedding by uniform manifold approximation and projection (UMAP) showed the separation of samples from different groups in reduced dimensions. The discriminant models were established by extreme gradient boosting (XGBoost) based on the features extracted by KPIC2. Results showed that features extracted by KPIC2 achieved 100% classification accuracy on the test sets of the liquor dataset and the LCC dataset, which demonstrated the rationality of the XGBoost model based on KPIC2 compared with the results of XCMS (92% and 96% for liquor and LCC datasets respectively). Finally, XGBoost can achieve better performance than the linear method and traditional nonlinear modeling methods on these datasets. UMAP and XGBoost are integrated into KPIC2 package to extend its performance in complex situations, which are not only able to effectively process nonlinear dataset but also can greatly improve the accuracy of data analysis in non-target metabolomics.

Comparing Epsilon Greedy and Thompson Sampling model for Multi-Armed Bandit algorithm on Marketing Dataset

A/B checking is a regular measure in many marketing procedures for e-Commerce companies. Through well-designed A/B research, advertisers can gain insight about when and how marketing efforts can be maximized and active promotions driven. Whilst many algorithms for the problem are theoretically well developed, empirical confirmation is typically restricted. In practical terms, standard A/B experimentation makes less money relative to more advanced machine learning methods. This paper presents a thorough empirical study of the most popular multi-strategy algorithms. Three important observations can be made from our results. First, simple heuristics such as Epsilon Greedy and Thompson Sampling outperform theoretically sound algorithms in most settings by a significant margin. In this report, the state of A/B testing is addressed, some typical A/B learning algorithms (Multi-Arms Bandits) used to optimize A/B testing are described and comparable. We found that Epsilon Greedy, be an exceptional winner to optimize payouts in this situation.

A novel systematic approach for cancer treatment prognosis and its applications in oropharyngeal cancer with microRNA biomarkers.

Predicting early in treatment whether a tumor is likely to respond to treatment is one of the most difficult yet important tasks in providing personalized cancer care. Most oropharyngeal squamous cell carcinoma (OPSCC) patients receive standard cancer therapy. However, the treatment outcomes vary significantly and are difficult to predict. Multiple studies indicate that microRNAs (miRNAs) are promising cancer biomarkers for the prognosis of oropharyngeal cancer. The reliable and efficient use of miRNAs for patient stratification and treatment outcome prognosis is still a very challenging task, mainly due to the relatively high dimensionality of miRNAs compared to the small number of observation sets; the redundancy, irrelevancy and uncertainty in the large amount of miRNAs; and the imbalanced observation patient samples. In this study, a new machine learning-based prognosis model was proposed to stratify subsets of OPSCC patients with low and high risks for treatment failure. The model cascaded a two-stage prognostic biomarker selection method and an evidential K-nearest neighbors classifier to address the challenges and improve the accuracy of patient stratification. The model has been evaluated on miRNA expression profiling of 150 oropharyngeal tumors by use of overall survival and disease-specific survival as the end points of disease treatment outcomes, respectively. The proposed method showed superior performance compared to other advanced machine-learning methods in terms of common performance quantification metrics. The proposed prognosis model can be employed as a supporting tool to identify patients who are likely to fail standard therapy and potentially benefit from alternative targeted treatments.

Sex Differences in Functional Topography of Association Networks

Prior work has shown that there is substantial interindividual variation in the spatial distribution of functional networks across the cerebral cortex, or functional topography. However, it remains unknown whether there are sex differences in the topography of individualized networks in youth. Here we leveraged an advanced machine learning method (sparsity-regularized non-negative matrix factorization) to define individualized functional networks in 693 youth (ages 8-23 years) who underwent functional magnetic resonance imaging as part of the Philadelphia Neurodevelopmental Cohort. Multivariate pattern analysis using support vector machines classified participant sex based on functional topography with 83% accuracy (p<0.0001). Brain regions most effective in classifying participant sex belonged to association networks, including the ventral attention and default mode networks. Mass-univariate analyses using generalized additive models with penalized splines provided convergent results. Comparative analysis using transcriptomic data from the Allen Human Brain Atlas revealed that sex differences in multivariate patterns of functional topography correlated with the expression of genes on the X-chromosome. These results identify normative developmental sex differences in the functional topography of association networks and highlight the role of sex as a biological variable in shaping brain development in youth.