Permutation Importance Research Articles

Comparative Study of AI-Enabled DDoS Detection Technologies in SDN

Software-defined networking (SDN) is becoming the standard for the management of networks due to its scalability and flexibility to program the network. SDN provides many advantages but it also involves some specific security problems; for example, the controller can be taken down using cyber attacks, which can result in the whole network shutting down, creating a single point of failure. In this paper, DDoS attacks in SDN are detected using AI-enabled machine and deep learning models with some specific features for a dataset under normal DDoS traffic. In our approach, the initial dataset is collected from 84 features on Kaggle and then the 20 top features are selected using a permutation importance algorithm. The dataset is learned and tested with five AI-enabled models. Our experimental results show that the use of a machine learning-based random forest model achieves the highest accuracy rate of 99.97% in DDoS attack detection in SDN. Our contributions through this study are, firstly, that we found the top 20 features that contributed to DDoS attacks. Secondly, we reduce the time and cost of comparing various learning models and their performance in determining a learning model suitable for DDoS detection. Finally, various experimental methods to evaluate the performance of the learning model are presented so that related researchers can utilize them.

Downscaling inversion of GRACE-derived groundwater storage changes based on ensemble learning

ABSTRACT Gravity Recovery and Climate Experiment (GRACE) satellite data monitors changes in terrestrial water storage, including groundwater, at a regional scale. However, the coarse spatial resolution limits its applicability to small watershed areas. This study introduces a novel ensemble learning-based model using meteorological and topographical data to enhance spatial resolution. The effectiveness was evaluated using groundwater-level observation data from the Henan rainstorm-affected area in July 2021. The factors influencing Groundwater Storage Anomalies (GWSA) were explored using Permutation Importance (PI) and other methods. The results demonstrate that feature engineering and Blender ensemble learning improve downscaling accuracy; the Root Mean Square Error (RMSE) can be reduced by up to 18.95%. Furthermore, Blender ensemble learning decreased the RMSE by 3.58%, achieving an R-Square (R2) value of 0.7924. Restricting the downscaling inversion to June–August data greatly enhanced the accuracy, as evidenced by a holdout dataset test with an R2 value of 0.8247. The overall GWSA variation from January to August exhibited ‘slow rise, slow fall, sharp fall, and sharp rise.’ Additionally, heavy rain exhibits a lag effect on the groundwater supply. Meteorological and topographical factors drive fluctuations in GWSA values and changes in spatial distribution. Human activities also have a significant impact.

Consistent and unbiased variable selection under indepedent features using Random Forest permutation importance

Variable selection in sparse regression models is an important task as applications ranging from biomedical research to econometrics have shown. Especially for higher dimensional regression problems, for which the regression function as the link between response and covariates cannot be directly detected, the selection of informative variables is challenging. Under these circumstances, the Random Forest method is an helpful tool to predict new outcomes while delivering measures for variable selection. One common approach is the usage of the permutation importance. Due to its intuitive idea and flexible usage, it is important to explore circumstances, for which the permutation importance based on Random Forest correctly indicates informative covariates. Regarding the latter, we deliver theoretical guarantees for the validity of the permutation importance measure under specific assumptions such as the mutual independence of the features and prove its (asymptotic) unbiasedness, while under slightly stricter assumptions, consistency of the permutation importance measure is established. An extensive simulation study supports our findings.

Flight departure delay forecasting

The most accurate methods to predict delays have become essential to mitigate their generation and proliferation (and prevent their spread to other airports in the network). Given that both the delay time dependence function and the input variables on which it depends are not fully known, the objective of this research is to predict the delays in the departure of scheduled commercial flights through a methodology that uses predictive tools based on machine learning/deep learning (ML/DL), with supervised training in regression, based on the available flight datasets. The novel contribution of this work is, first, to make a comparison of the predictions in terms of means and statistical variance of the different ML/DL models implemented (ten in total) and, secondly, to determine the coefficients of the importance of the features or flight attributes. Using ML methods known as permutation importance, it is possible to rank the importance of flight attributes by their influence in determining the delay time and reduce the problem of selecting the most important flight attributes. The data for the analysis was obtained from the Colombian airport system for the year 2018. From the results obtained, it is worth mentioning that the model that presents the best performance is the Ensemble, or combination method of Random Forest Regressor models, configured with 2,000 trees within the forest, with an acceptable prediction range measured with the root-mean-square-error (RMSE) metric of 16 to 33 minutes (prediction of time of the flight departure delay) depending on the scenario analysed.

An Approach Based on Recurrent Neural Networks and Interactive Visualization to Improve Explainability in AI Systems

This paper investigated the importance of explainability in artificial intelligence models and its application in the context of prediction in Formula (1). A step-by-step analysis was carried out, including collecting and preparing data from previous races, training an AI model to make predictions, and applying explainability techniques in the said model. Two approaches were used: the attention technique, which allowed visualizing the most relevant parts of the input data using heat maps, and the permutation importance technique, which evaluated the relative importance of features. The results revealed that feature length and qualifying performance are crucial variables for position predictions in Formula (1). These findings highlight the relevance of explainability in AI models, not only in Formula (1) but also in other fields and sectors, by ensuring fairness, transparency, and accountability in AI-based decision making. The results highlight the importance of considering explainability in AI models and provide a practical methodology for its implementation in Formula (1) and other domains.

How Important Is Satellite-Retrieved Aerosol Optical Depth in Deriving Surface PM2.5 Using Machine Learning?

PM2.5 refers to the total mass concentration of tiny particulates in the atmosphere near the surface, obtained by means of in situ observations and satellite remote sensing. Given the highly limited number of ground observation stations of inhomogeneous distribution and an ill-posed remote sensing approach, increasing efforts have been devoted to the application of machine-learning (ML) models to both ground and satellite data. A key satellite-derived parameter, aerosol optical thickness (AOD), has been most commonly used as a proxy of PM2.5, although their correlation is fraught with large uncertainties. A critical question that has been overlooked concerns how much AOD helps to improve the retrieval of PM2.5 relative to its uncertainty incurred concurrently. The question is addressed here by taking advantage of high-density PM2.5 stations in eastern China to evaluate the contributions of AOD, determined as the difference in the accuracy of PM2.5 retrievals with and without AOD for varying densities of PM2.5 stations, using four popular ML models (i.e., Random Forest, Extra-trees, XGBoost, and LightGBM). Our results reveal that as the density of monitoring stations decreases, both the feature importance and permutation importance of satellite AOD demonstrate a consistent upward trend (p < 0.05). Furthermore, the ML models without AOD exhibit faster declines in overall accuracy and predictive ability compared with the models with AOD assessed using the sample-based and station-based (spatial) independent cross-validation approaches. Overall, a 10% reduction in the number of stations results in an increase of 0.7–1.2% and 0.6–1.2% in uncertainty in estimated and predicted accuracies, respectively. These findings attest to the indispensable role of satellite AOD in the PM2.5 retrieval process through ML because it can significantly mitigate the negative impact of the sparse distribution of monitoring sites. This role becomes more important as the number of PM2.5 stations decreases.

Machine Learning Identifies a Signature of Nine Exosomal RNAs That Predicts Hepatocellular Carcinoma.

Hepatocellular carcinoma (HCC) is the third leading cause of cancer-related death worldwide. Although alpha fetoprotein (AFP) remains a commonly used serological marker of HCC, the sensitivity and specificity of AFP in detecting HCC is often limited. Exosomal RNA has emerged as a promising diagnostic tool for various cancers, but its use in HCC detection has yet to be fully explored. Here, we employed Machine Learning on 114,602 exosomal RNAs to identify a signature that can predict HCC. The exosomal expression data of 118 HCC patients and 112 healthy individuals were stratified split into Training, Validation and Unseen Test datasets. Feature selection was then performed on the initial training dataset using permutation importance, and the predictive performance of the selected features were tested on the validation dataset using Support Vector Machine (SVM) Classifier. A minimum of nine features were identified to be predictive of HCC and these nine features were then evaluated across six different models in an unseen test set. These features, mainly in the immune, platelet/neutrophil and cytoskeletal pathways, exhibited good predictive performance with ROC-AUC from 0.79-0.88 in the unseen test set. Hence, these nine exosomal RNAs have potential to be clinically useful minimally invasive biomarkers for HCC.

Modelling point mass balance for the glaciers of the Central European Alps using machine learning techniques

Abstract. Glacier mass balance is typically estimated using a range of in situ measurements, remote sensing measurements, and physical and temperature index modelling techniques. With improved data collection and access to large datasets, data-driven techniques have recently gained prominence in modelling natural processes. The most common data-driven techniques used today are linear regression models and, to some extent, non-linear machine learning models such as artificial neural networks. However, the entire host of capabilities of machine learning modelling has not been applied to glacier mass balance modelling. This study used monthly meteorological data from ERA5-Land to drive four machine learning models: random forest (ensemble tree type), gradient-boosted regressor (ensemble tree type), support vector machine (kernel type), and artificial neural networks (neural type). We also use ordinary least squares linear regression as a baseline model against which to compare the performance of the machine learning models. Further, we assess the requirement of data for each of the models and the requirement for hyperparameter tuning. Finally, the importance of each meteorological variable in the mass balance estimation for each of the models is estimated using permutation importance. All machine learning models outperform the linear regression model. The neural network model depicted a low bias, suggesting the possibility of enhanced results in the event of biased input data. However, the ensemble tree-based models, random forest and gradient-boosted regressor, outperformed all other models in terms of the evaluation metrics and interpretability of the meteorological variables. The gradient-boosted regression model depicted the best coefficient of determination value of 0.713 and a root mean squared error of 1.071 m w.e. The feature importance values associated with all machine learning models suggested a high importance of meteorological variables associated with ablation. This is in line with predominantly negative mass balance observations. We conclude that machine learning techniques are promising in estimating glacier mass balance and can incorporate information from more significant meteorological variables as opposed to a simplified set of variables used in temperature index models.

Sex matters: acute functional connectivity changes as markers of remission in late-life depression differ by sex.

The efficacy of antidepressant treatment in late-life is modest, a problem magnified by an aging population and increased prevalence of depression. Understanding the neurobiological mechanisms of treatment response in late-life depression (LLD) is imperative. Despite established sex differences in depression and neural circuits, sex differences associated with fMRI markers of antidepressant treatment response are underexplored. In this analysis, we assess the role of sex on the relationship of acute functional connectivity changes with treatment response in LLD. Resting state fMRI scans were collected at baseline and day one of SSRI/SNRI treatment for 80 LLD participants. One-day changes in functional connectivity (differential connectivity) were related to remission status after 12 weeks. Sex differences in differential connectivity profiles that distinguished remitters from non-remitters were assessed. A random forest classifier was used to predict the remission status with models containing various combinations of demographic, clinical, symptomatological, and connectivity measures. Model performance was assessed with area under the curve, and variable importance was assessed with permutation importance. The differential connectivity profile associated with remission status differed significantly by sex. We observed evidence for a difference in one-day connectivity changes between remitters and non-remitters in males but not females. Additionally, prediction of remission was significantly improved in male-only and female-only models over pooled models. Predictions of treatment outcome based on early changes in functional connectivity show marked differences between sexes and should be considered in future MR-based treatment decision-making algorithms.

Selection of single machine learning model for designing compressive strength of stabilized soil containing lime, cement and bitumen

The unconfined compressive strength (Qu) is one of the most important criteria of stabilized soil to design in order to evaluate the effective of soft soil improvement. The unconfined compressive strength of stabilized soil is strongly affected by numerous factors such as the soil properties, the binder content, etc. Machine Learning (ML) approach can take into account these factors to predict the unconfined compressive strength (Qu) with high performance and reliability. The aim of this paper is to select a single ML model to design Qu of stabilized soil containing some chemical stabilizer agents such as lime, cement and bitumen. In order to build the single ML model, a database is created based on the literature investigation. The database contains 200 data samples, 12 input variables (Liquid limit, Plastic limit, Plasticity index, Linear shrinkage, Clay content, Sand content, Gravel content, Optimum water content, Density of stabilized soil, Lime content, Cement content, Bitumen content) and the output variable Qu. The performance and reliability of ML model are evaluated by the popular validation technique Monte Carlo simulation with aided of three criteria metrics including coefficient of determination R2, Root Mean Square Error (RMSE) and Mean Square Error (MAE). ML model based on Gradient Boosting algorithm is selected as highest performance and highest reliability ML model for designing Qu of stabilized soil. Explanation of feature effects on the unconfined compressive strength Qu of stabilized soil is carried out by Permutation importance, Partial Dependence Plot (PDP 2D) in two dimensions and SHapley Additive exPlanations (SHAP) local value. The ML model proposed in this investigation is single and useful for professional engineers with using the mapping Maximal dry density-Linear shrinkage created by PDP 2D.

Discovering epistasis interactions in Alzheimer’s disease using integrated framework of ensemble learning and multifactor dimensionality reduction (MDR)

Alzheimer's disease (AD) is a complex disorder with strong genetic factors. The proposed framework is applied to Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset. We present a novel framework integrating ensemble learning and MDR constructive induction algorithm to discover epistasis interactions associated with AD in a computationally efficient method. Discovering epistasis interactions is a big challenge and significantly impacts personalized medicine (PM). The applied ensemble learning algorithms are random forests (RF) with Gini index and permutation importance, Extreme Gradient Boosting (XGBoost), and classification and regression trees (CART). The classification accuracy of 5-way models varied between (0.8674–0.8758), whereas the accuracy of 2-way, 3-way, and 4-way models varied between (0.6515–0.6649), (0.7071–0.7170), and (0.7811–0.7878) respectively. The promising results of this proposed framework show high-ranked risk genes and up to 5-way epistasis models that contribute to the disease risk efficiently and at higher accuracy.

Neural Network Approach to Investigating the Importance of Test Items for Predicting Physical Activity in Chronic Obstructive Pulmonary Disease.

Contracting COPD reduces a patient's physical activity and restricts everyday activities (physical activity disorder). However, the fundamental cause of physical activity disorder has not been found. In addition, costly and specialized equipment is required to accurately examine the disorder; hence, it is not regularly assessed in normal clinical practice. In this study, we constructed a machine learning model to predict physical activity using test items collected during the normal care of COPD patients. In detail, we first applied three types of data preprocessing methods (zero-padding, multiple imputation by chained equations (MICE), and k-nearest neighbor (kNN)) to complement missing values in the dataset. Then, we constructed several types of neural networks to predict physical activity. Finally, permutation importance was calculated to identify the importance of the test items for prediction. Multifactorial analysis using machine learning, including blood, lung function, walking, and chest imaging tests, was the unique point of this research. From the experimental results, it was found that the missing value processing using MICE contributed to the best prediction accuracy (73.00%) compared to that using zero-padding (68.44%) or kNN (71.52%), and showed better accuracy than XGBoost (66.12%) with a significant difference (p < 0.05). For patients with severe physical activity reduction (total exercise < 1.5), a high sensitivity (89.36%) was obtained. The permutation importance showed that "sex, the number of cigarettes, age, and the whole body phase angle (nutritional status)" were the most important items for this prediction. Furthermore, we found that a smaller number of test items could be used in ordinary clinical practice for the screening of physical activity disorder.

Machine learning-based pedotransfer functions to predict soil water characteristics curves

Soil water characteristic curve (SWCC) is a key property in characterizing unsaturated soil behaviors. Despite considerable progress in predicting methods, predicting SWCCs remains challenging owing to their huge uncertainty. This study exploited the advantages of seven machine learning (ML) models and the unsaturated soil database (UNSODA) to develop a new pedotransfer function (PTF) for estimating SWCC. The importance of UNSODA attributes, including pressure head, soil textural information, state parameters, and particle density, was evaluated using permutation importance and Shapley values. In addition, the performance of ML-PTFs for seven feature selection scenarios was measured based on the evaluated rank of feature importance using Shapley values. The PTF implemented on the extreme gradient boosting (XGB) model yielded the best performance with the highest coefficient of determination of 0.972, which is comparable to the performance documented in the literature. In addition, the pressure head was evaluated as the most important feature, followed by sand fraction, clay fraction, and bulk density. Noticeably, the performance of the seven ML-PTFs converged when the number of features was greater than four (the four most important features), indicating the possibility of excluding silt fraction, particle density, and porosity in developing ML-PTF to predict SWCCs. Finally, to manifest the practical applications the developed XGB-PTF was integrated into the Bayesian optimization to approximate the matric suction profile in Ho Chi Minh City.

Towards interpreting machine learning models for predicting soil moisture droughts

Determination of the dominant factors which affect soil moisture (SM) predictions for drought analysis is an essential step to assess the reliability of the prediction results. However, artificial intelligence (AI) based drought modelling only provides prediction results without the physical interpretation of the models. Here, we propose an explainable AI (XAI) framework to reveal the modelling of SM drought events. Random forest based site-specific SM prediction models were developed using the data from 30 sites, covering 8 vegetation types. The unity of multiply XAI tools was applied to interpret the site-models both globally (generally) and locally. Globally, the models were interpreted using two methods: permutation importance and accumulated local effect (ALE). On the other hand, for each drought event, the models were interpreted locally via Shapley additive explanations (SHAP), local interpretable model-agnostic explanation (LIME) and individual conditional expectation (ICE) methods. Globally, the dominant features for SM predictions were identified as soil temperature, atmospheric aridity, time variables and latent heat flux. But through local interpretations of the drought events, SM showed a greater reliance on soil temperature, atmospheric aridity and latent heat flux at grass sites, with higher correlation to the time-dependent parameters at the sites located in forests. The temporal variation of the feature which effects the drought events was also demonstrated. The interpretation could shed light on how predictions are made and could promote the application of AI techniques in drought prediction, which may be useful for irrigation and water resource management.

Hybrid BO-XGBoost and BO-RF Models for the Strength Prediction of Self-Compacting Mortars with Parametric Analysis.

Self-compacting mortar (SCM) has superior workability and long-term durable performance compared to traditional mortar. The strength of SCM, including both its compressive and flexural strengths, is a crucial property that is determined by appropriate curing conditions and mix design parameters. In the context of materials science, predicting the strength of SCM is challenging because of multiple influencing factors. This study employed machine learning techniques to establish SCM strength prediction models. Based on ten different input parameters, the strength of SCM specimens were predicted using two different types of hybrid machine learning (HML) models, namely Extreme Gradient Boosting (XGBoost) and the Random Forest (RF) algorithm. HML models were trained and tested by experimental data from 320 test specimens. In addition, the Bayesian optimization method was utilized to fine tune the hyperparameters of the employed algorithms, and cross-validation was employed to partition the database into multiple folds for a more thorough exploration of the hyperparameter space while providing a more accurate assessment of the model's predictive power. The results show that both HML models can successfully predict the SCM strength values with high accuracy, and the Bo-XGB model demonstrated higher accuracy (R2 = 0.96 for training and R2 = 0.91 for testing phases) for predicting flexural strength with low error. In terms of compressive strength prediction, the employed BO-RF model performed very well, with R2 = 0.96 for train and R2 = 0.88 testing stages with minor errors. Moreover, the SHAP algorithm, permutation importance and leave-one-out importance score were used for sensitivity analysis to explain the prediction process and interpret the governing input variable parameters of the proposed HML models. Finally, the outcomes of this study might be applied to guide the future mix design of SCM specimens.

Interpretable Battery Cycle Life Range Prediction Using Early Cell Degradation Data

Battery cycle life prediction using early degradation data has many potential applications throughout the battery product life cycle. For that reason, various data-driven methods have been proposed for point prediction of battery cycle life with minimum knowledge of the battery degradation mechanisms. However, managing the rapidly increasing amounts of batteries at end-of-life with lower economic and technical risk requires prediction of cycle life with quantified uncertainty, which is still lacking. The interpretability (i.e., the reason for high prediction accuracy) of these advanced data-driven methods is also worthy of investigation. Here, a physics-informed Quantile Regression Forest (QRF) model, having the advantage of not assuming any specific distribution of cycle life, is introduced to make cycle life range prediction with uncertainty quantified as the width of the prediction interval, in addition to point predictions with high accuracy. The hyperparameters of the QRF model are optimized with a proposed alpha-logistic-weighted criterion so that the coverage probabilities associated with the prediction intervals are calibrated. The interpretability of the final QRF model is explored with two global model-agnostic methods, namely permutation importance and partial dependence plot.

Interpretable machine learning model to predict survival days of malignant brain tumor patients

An artificial intelligence (AI) model’s performance is strongly influenced by the input features. Therefore, it is vital to find the optimal feature set. It is more crucial for the survival prediction of the glioblastoma multiforme (GBM) type of brain tumor. In this study, we identify the best feature set for predicting the survival days (SD) of GBM patients that outrank the current state-of-the-art methodologies. The proposed approach is an end-to-end AI model. This model first segments tumors from healthy brain parts in patients’ MRI images, extracts features from the segmented results, performs feature selection, and makes predictions about patients’ survival days (SD) based on selected features. The extracted features are primarily shape-based, location-based, and radiomics-based features. Additionally, patient metadata is also included as a feature. The selection methods include recursive feature elimination, permutation importance (PI), and finding the correlation between the features. Finally, we examined features’ behavior at local (single sample) and global (all the samples) levels. In this study, we find that out of 1265 extracted features, only 29 dominant features play a crucial role in predicting patients’ SD. Among these 29 features, one is metadata (age of patient), three are location-based, and the rest are radiomics features. Furthermore, we find explanations of these features using post-hoc interpretability methods to validate the model’s robust prediction and understand its decision. Finally, we analyzed the behavioral impact of the top six features on survival prediction, and the findings drawn from the explanations were coherent with the medical domain. We find that after the age of 50 years, the likelihood of survival of a patient deteriorates, and survival after 80 years is scarce. Again, for location-based features, the SD is less if the tumor location is in the central or back part of the brain. All these trends derived from the developed AI model are in sync with medically proven facts. The results show an overall 33% improvement in the accuracy of SD prediction compared to the top-performing methods of the BraTS-2020 challenge.

Lithology identification based on interpretability integration learning

An interpretability model for intelligent lithology identification is proposed, which utilizes Ensemble Learning Stacking, Permutation Importance (PI), and Local Interpretable Model-agnostic Explanations (LIME) techniques. The aim of this method is to provide more accurate geological information and scientific support for oil and gas resource exploration. Two logging datasets from the public domain were used as experiments, and support vector machine (SVM), random forest (RF), and naive Bayes (NB) were employed as base learners, while SVM was utilized as the meta learner for lithology classification via stacking algorithm. The accuracy of the model was verified using evaluation metrics such as Area Under Curve (AUC), precision, recall, and F1-score. The PI and LIME techniques were employed to explain the lithology identification model. The results indicate that the stacking algorithm produced the best indexes and highest prediction accuracy. With respect to overall interpretation, PHIND, GR, and RT were found to have the most significant influence on lithology identification in a natural gas protection area in the United States, while DEN, CAL, and PEF were observed to be the most influential variables for lithology identification in the Daqing Oilfield in China. From the perspective of a single sample, the LIME algorithm can provide a quantitative prediction probability and degree of influence of the characteristic variables.

0621 A machine learning model to predict the risk of perinatal depression from sleep data in healthy pregnant women

Abstract Introduction Perinatal depression (PND) is a complication of pregnancy associated with serious health consequences for both mothers and their babies. Identifying predisposing factors for PND during early pregnancy is key to early detection of women who may be at greater risk of developing this condition. Machine Learning (ML) techniques have recently been applied in a few studies, mostly to predict postpartum depression. None of them, however, considered sleep data in model building, neither focused on PND. Methods By analyzing data from a multicenter, prospective cohort study on sleep and mood changes during the perinatal period (the “Life-ON” study), we constructed a ML model for PND risk prediction and tested it in a cross-validation setting. PND was assessed using an EPDS score &amp;gt;12 during 9 visits from early pregnancy until 6 months postpartum. In a bivariate analysis, 47 sociodemographic, psychological, blood-based, medical/gynecological, and subjective sleep variables, collected from 439 pregnant women (33.7±4.2 yrs.) during the first trimester of gestation, as well as 33 polysomnographic (PSG) parameters, recorded from 353 women (33.6±4.2 yrs.) during the second trimester, were correlated with PND. A support vector machine (SVM) model was then trained using the 10-features with the highest permutation importance to predict the individual risk of PND for each woman. Results Among all variables considered, sleep quality (PSQI) and insomnia symptoms (ISI) (p&amp;lt; 0.001), daytime sleepiness (ESS) and RLS severity (IRLS) (p&amp;lt; 0.05), as well as the PSG parameters Apnea Index (p=0.001), number of central hypopneas and percentage of sleep stage N2 (p&amp;lt; 0.05), were all positively correlated to PND. The PSQI and ISI scores were also selected by the SVM classifier, which achieved a mean AUROC of 0.777 and an AUPRC of 0.393, corresponding to a sensitivity of 54.3% and a specificity of 82.6% in identifying women at risk for PND. Conclusion In our data-driven ML model to predict the risk of PND during early pregnancy, subjective poor sleep quality and insomnia symptoms identified women at greater risk of developing PND, while none of the PSG variables improved model performance. Support (if any) Swiss National Science Foundation (grant: 320030_160250/1). The Italian Ministry of Health and Emilia-Romagna Region (grant: PE-2011-02348727).

Super Typhoon Rai’s Impacts on Siargao Tourism: Deciphering Tourists’ Revisit Intentions through Machine-Learning Algorithms

Super Typhoon Rai damaged Siargao’s tourism industry. Despite the reconstruction projects, there was still evidence of limited resources, destructed infrastructures, and destroyed natural resources. Therefore, this study aimed to examine the significant factors influencing tourists’ intentions to revisit Siargao after Super Typhoon Rai using feature selection, logistic regression (LR), and an artificial neural network (ANN). It employed three feature-selection techniques, namely, the filter method’s permutation importance (PI), the wrapper method’s Recursive Feature Elimination (RFE), and the embedded method’s Least Absolute Shrinkage and Selection Operator (LASSO). Each feature-selection technique was integrated into LR and the ANN. LASSO-ANN, with a 97.8146% model accuracy, was found to be the best machine-learning algorithm. The LASSO model performed at its best with a 0.0007 LASSO alpha value, resulting in 35 subfeatures and 8 primary features. LASSO subsets underwent the ANN model procedure, and the optimal parameter combination was 70% training size, 30% testing size, 30 hidden-layer nodes, tanh hidden-layer activation, sigmoid output-layer activation, and Adam optimization. All eight features were found to be significant. Among them, hedonic motivation and awareness of Typhoon Rai’s impact were considered the top-tier post-typhoon tourism factors, as they maintained at least 97% prediction accuracy. The findings could be elaborated by combining feature-selection techniques, utilizing demographic characteristics, assessing Siargao’s tourism before the typhoon, and expanding the context and participant selection. Nevertheless, none of the existing studies explored the combination of feature selection, LR, and ANNs in a post-typhoon tourism context. These unique methods and significant findings represent the study’s novelty. Furthermore, practical contributions were provided through economic resolutions focusing on tourism activities and communication revamping by the government, media outlets, and transportation companies.