The study of under- and over-sampling methods’ utility in analysis of highly imbalanced data on osteoporosis

Automated detection of selected tea leaf diseases in Bangladesh with convolutional neural network

Globally, tea production and its quality fundamentally depend on tea leaves, which are susceptible to invasion by pathogenic organisms. Precise and early-stage identification of plant foliage diseases is a key element in preventing and controlling the spreading of diseases that hinder yield and quality. Image processing techniques are a sophisticated tool that is rapidly gaining traction in the agricultural sector for the detection of a wide range of diseases with excellent accuracy. This study focuses on a pragmatic approach for automatically detecting selected tea foliage diseases based on convolutional neural network (CNN). A large dataset of 3330 images has been created by collecting samples from different regions of Sylhet division, the tea capital of Bangladesh. The proposed CNN model is developed based on tea leaves affected by red rust, brown blight, grey blight, and healthy leaves. Afterward, the model's prediction was validated with laboratory tests that included microbial culture media and microscopic analysis. The accuracy of this model was found to be 96.65%. Chiefly, the proposed model was developed in the context of the Bangladesh tea industry.

A feasibility study assessing quantitative indocyanine green angiographic predictors of reconstructive complications following nipple-sparing mastectomy

Introduction

Immediate post-mastectomy breast reconstruction offers benefits; however, complications can compromise outcomes. Intraoperative indocyanine green fluorescence angiography (ICGFA) may mitigate perfusion-related complications (PRC); however, its interpretation remains subjective. Here, we examine and develop methods for ICGFA quantification, including machine learning (ML) algorithms for predicting complications.

Methods

ICGFA video recordings of flap perfusion from a previous study of patients undergoing nipple-sparing mastectomy (NSM) with either immediate or staged immediate (delayed by a week due to perfusion insufficiency) reconstructions were analysed. Fluorescence intensity time series data were extracted, and perfusion parameters were interrogated for overall/regional associations with postoperative PRC. A naïve Bayes ML model was subsequently trained on a balanced data subset to predict PRC from the extracted meta-data.

Results

The analysable video dataset of 157 ICGFA featured females (average age 48 years) having oncological/risk-reducing NSM with either immediate (n=90) or staged immediate (n=26) reconstruction. For those delayed, peak brightness at initial ICGFA was lower (p<0.001) and significantly improved (both quicker-onset and brighter p=0.001) one week later. The overall PRC rate in reconstructed patients (n=116) was 11.2%, with such patients demonstrating significantly dimmer (overall, p=0.018, centrally, p=0.03, and medially, p=0.04) and slower-onset (p=0.039) fluorescent peaks with shallower slopes (p=0.012) than uncomplicated patients with ICGFA. Importantly, such relevant parameters were converted into a whole field of view heatmap potentially suitable for intraoperative display. ML predicted PRC with 84.6% sensitivity and 76.9% specificity.

Conclusion

Whole breast quantitative ICGFA assessment reveals statistical associations with PRC that are potentially exploitable via ML.

Multivariable time series classification for clinical mastitis detection and prediction in automated milking systems

In this study, we developed a machine learning framework to detect clinical mastitis (CM) at the current milking (i.e., the same milking) and predict CM at the next milking (i.e., one milking before CM occurrence) at the quarter level. Time series quarter-level milking data were extracted from an automated milking system (AMS). For both CM detection and prediction, the best classification performance was obtained from the decision tree-based ensemble models. Moreover, applying models on a data set containing data from the current milking and past 9 milkings before the current milking showed the best accuracy for detecting CM; modeling with a data set containing data from the current milking and past 7 milkings before the current milking yielded the best results for predicting CM. The models combined with oversampling methods resulted in specificity of 95 and 93% for CM detection and prediction, respectively, with the same sensitivity (82%) for both scenarios; when lowering specificity to 80 to 83%, undersampling techniques facilitated models to increase sensitivity to 95%. We propose a feasible machine learning framework to identify CM in a timely manner using imbalanced data from an AMS, which could provide useful information for farmers to manage the negative effects of CM.

DE-INTERACT: A machine-learning-based predictive tool for the drug-excipient interaction study during product development-Validation through Paracetamol and Vanillin as a case study

The compatibility of drugs with excipients plays a crucial role in the prospective stability of pharmaceutical formulations. Apart from real-time stability studies, conventional analytical tools like DSC, FTIR, NMR, and chromatography help identify the possibilities of drug-excipient interactions. Machine learning can assist in developing a predictive tool for drug-excipient incompatibility. In the present work, PubChem Fingerprint is employed as the descriptor of compounds that thoroughly represents the drug's and excipient's chemistry. The 881-bit binary fingerprints of each drug and excipient make 1762 inputs, and one categorical output makes an instance in the dataset. A dataset of more than 3500 instances of drugs and excipients is carefully selected from peer-reviewed research papers. Rigorous training of the Artificial Neural Network (ANN) model was performed with maximum validation accuracy, minimum validation loss, and maximum validation precision as the checkpoints. The machine learning model (DE-Interact) was trained, achieving training and validation accuracies of 0.9930 and 0.9161, respectively. The performance of the DE-Interact model was evaluated by confirming three incompatible predictions by conventional analytical tools. Paracetamol with vanillin, paracetamol with methylparaben, and brinzolamide with polyethyleneglycol are these instances which are predicted as incompatible by the DE-Interact. DSC, FTIR, HPTLC, and HPLC analysis confirm the prediction. The present work offers a reliable DE-Interact tool for quick referencing while selecting excipients in formulation design.

Comparison of mortality prediction models for road traffic accidents: an ensemble technique for imbalanced data

BACKGROUND

Injuries caused by RTA are classified under the International Classification of Diseases-10 as 'S00-T99' and represent imbalanced samples with a mortality rate of only 1.2% among all RTA victims. To predict the characteristics of external causes of road traffic accident (RTA) injuries and mortality, we compared performances based on differences in the correction and classification techniques for imbalanced samples.

METHODS

The present study extracted and utilized data spanning over a 5-year period (2013-2017) from the Korean National Hospital Discharge In-depth Injury Survey (KNHDS), a national level survey conducted by the Korea Disease Control and Prevention Agency, A total of eight variables were used in the prediction, including patient, accident, and injury/disease characteristics. As the data was imbalanced, a sample consisting of only severe injuries was constructed and compared against the total sample. Considering the characteristics of the samples, preprocessing was performed in the study. The samples were standardized first, considering that they contained many variables with different units. Among the ensemble techniques for classification, the present study utilized Random Forest, Extra-Trees, and XGBoost. Four different over- and under-sampling techniques were used to compare the performance of algorithms using "accuracy", "precision", "recall", "F1", and "MCC".

RESULTS

The results showed that among the prediction techniques, XGBoost had the best performance. While the synthetic minority oversampling technique (SMOTE), a type of over-sampling, also demonstrated a certain level of performance, under-sampling was the most superior. Overall, prediction by the XGBoost model with samples using SMOTE produced the best results.

CONCLUSION

This study presented the results of an empirical comparison of the validity of sampling techniques and classification algorithms that affect the accuracy of imbalanced samples by combining two techniques. The findings could be used as reference data in classification analyses of imbalanced data in the medical field.

High-Frequency Direction Forecasting of the Futures Market Using a Machine-Learning-Based Method

Futures price-movement-direction forecasting has always been a significant and challenging subject in the financial market. In this paper, we propose a combination approach that integrates the XGBoost (eXtreme Gradient Boosting), SMOTE (Synthetic Minority Oversampling Technique), and NSGA-II (Non-dominated Sorting Genetic Algorithm-II) methods. We applied the proposed approach on the direction prediction and simulation trading of rebar futures, which are traded on the Shanghai Futures Exchange. Firstly, the minority classes of the high-frequency rebar futures price change magnitudes are oversampled using the SMOTE algorithm to overcome the imbalance problem of the class data. Then, XGBoost is adopted to construct a multiclassification model for the price-movement-direction prediction. Next, the proposed approach employs NSGA-II to optimize the parameters of the pre-designed trading rule for trading simulation. Finally, the price-movement direction is predicted, and we conducted the high-frequency trading based on the optimized XGBoost model and the trading rule, with the classification and trading performances empirically evaluated by four metrics over four testing periods. Meanwhile, the LIME (Local Interpretable Model-agnostic Explanations) is applied as a model explanation approach to quantify the prediction contributions of features to the forecasting samples. From the experimental results, we found that the proposed approach performed best in terms of direction prediction accuracy, profitability, and return–risk ratio. The proposed approach could be beneficial for decision-making of the rebar traders and related companies engaged in rebar futures trading.

Clustering-based adaptive data augmentation for class-imbalance in machine learning (CADA): additive manufacturing use case

Large amount of data are generated from in-situ monitoring of additive manufacturing (AM) processes which is later used in prediction modelling for defect classification to speed up quality inspection of products. A high volume of this process data is defect-free (majority class) and a lower volume of this data has defects (minority class) which result in the class-imbalance issue. Using imbalanced datasets, classifiers often provide sub-optimal classification results, i.e. better performance on the majority class than the minority class. However, it is important for process engineers that models classify defects more accurately than the class with no defects since this is crucial for quality inspection. Hence, we address the class-imbalance issue in manufacturing process data to support in-situ quality control of additive manufactured components. For this, we propose cluster-based adaptive data augmentation (CADA) for oversampling to address the class-imbalance problem. Quantitative experiments are conducted to evaluate the performance of the proposed method and to compare with other selected oversampling methods using AM datasets from an aerospace industry and a publicly available casting manufacturing dataset. The results show that CADA outperformed random oversampling and the SMOTE method and is similar to random data augmentation and cluster-based oversampling. Furthermore, the results of the statistical significance test show that there is a significant difference between the studied methods. As such, the CADA method can be considered as an alternative method for oversampling to improve the performance of models on the minority class.

ASN-SMOTE: a synthetic minority oversampling method with adaptive qualified synthesizer selection

Oversampling is a promising preprocessing technique for imbalanced datasets which generates new minority instances to balance the dataset. However, improper generated minority instances, i.e., noise instances, may interfere the learning of the classifier and impact it negatively. Given this, in this paper, we propose a simple and effective oversampling approach known as ASN-SMOTE based on the k-nearest neighbors and the synthetic minority oversampling technology (SMOTE). ASN-SMOTE first filters noise in the minority class by determining whether the nearest neighbor of each minority instance belongs to the minority or majority class. After that, ASN-SMOTE uses the nearest majority instance of each minority instance to effectively perceive the decision boundary, inside which the qualified minority instances are selected adaptively for each minority instance by the proposed adaptive neighbor selection scheme to synthesize new minority instance. To substantiate the effectiveness, ASN-SMOTE has been applied to three different classifiers and comprehensive experiments have been conducted on 24 imbalanced benchmark datasets. ASN-SMOTE is also extensively compared with nine notable oversampling algorithms. The results show that ASN-SMOTE achieves the best results in the majority of datasets. The ASN-SMOTE implementation is available at: https://www.github.com/yixinkai123/ASN-SMOTE/.

Two novelty learning models developed based on deep cascade forest to address the environmental imbalanced issues: A case study of drinking water quality prediction

Environmental quality data sets are typically imbalanced, because environmental pollution events are rarely observed in daily life. Prediction of imbalanced data sets is a major challenge in machine learning. Our recent work has shown deep cascade forest (DCF), as a base learning model, is promising to be recommended for environmental quality prediction. Although some traditional models were improved by introducing the cost matrix, little is known about whether cost matrix could enhance the prediction performance of DCF. Additionally, feature extraction is also an important way to potentially improve the model's ability to predict the imbalanced data. Here, we developed two novelty learning models based on DCF: cost-sensitive DCF (CS-DCF) and DCF that combines unsupervised learning models and greedy methods (USM-DCF-G). Subsequently, CS-DCF and USM-DCF-G were successfully verified by an imbalanced drinking water quality data set. Our data presented both CS-DCF and USM-DCF-G show better prediction performance than that of DCF alone did. In particular, USM-DCF-G shows the best performance with the highest F1-score (95.12 ± 2.56%), after feature extraction and selection by using unsupervised learning models and greedy methods. Thus, the two learning models, especially USM-DCF-G, were promising learning models to address environmental imbalanced issues and accurately predict environmental quality.

Limited progress in nutrient pollution in the U.S. caused by spatially persistent nutrient sources

Human agriculture, wastewater, and use of fossil fuels have saturated ecosystems with nitrogen and phosphorus, threatening biodiversity and human water security at a global scale. Despite efforts to reduce nutrient pollution, carbon and nutrient concentrations have increased or remained high in many regions. Here, we applied a new ecohydrological framework to ~12,000 water samples collected by the U.S. Environmental Protection Agency from streams and lakes across the contiguous U.S. to identify spatial and temporal patterns in nutrient concentrations and leverage (an indicator of flux). For the contiguous U.S. and within ecoregions, we quantified trends for sites sampled repeatedly from 2000 to 2019, the persistence of spatial patterns over that period, and the patch size of nutrient sources and sinks. While we observed various temporal trends across ecoregions, the spatial patterns of nutrient and carbon concentrations in streams were persistent across and within ecoregions, potentially because of historical nutrient legacies, consistent nutrient sources, and inherent differences in nutrient removal capacity for various ecosystems. Watersheds showed strong critical source area dynamics in that 2-8% of the land area accounted for 75% of the estimated flux. Variability in nutrient contribution was greatest in catchments smaller than 250 km2 for most parameters. An ensemble of four machine learning models confirmed previously observed relationships between nutrient concentrations and a combination of land use and land cover, demonstrating how human activity and inherent nutrient removal capacity interactively determine nutrient balance. These findings suggest that targeted nutrient interventions in a small portion of the landscape could substantially improve water quality at continental scales. We recommend a dual approach of first prioritizing the reduction of nutrient inputs in catchments that exert disproportionate influence on downstream water chemistry, and second, enhancing nutrient removal capacity by restoring hydrological connectivity both laterally and vertically in stream networks.

Transformer-based CNNs: Mining Temporal Context Information for Multi-sound COVID-19 Diagnosis

Due to the COronaVIrus Disease 2019 (COVID-19) pandemic, early screening of COVID-19 is essential to prevent its transmission. Detecting COVID-19 with computer audition techniques has in recent studies shown the potential to achieve a fast, cheap, and ecologically friendly diagnosis. Respiratory sounds and speech may contain rich and complementary information about COVID-19 clinical conditions. Therefore, we propose training three deep neural networks on three types of sounds (breathing/counting/vowel) and assembling these models to improve the performance. More specifically, we employ Convolutional Neural Networks (CNNs) to extract spatial representations from log Mel spectrograms and a multi-head attention mechanism in the transformer to mine temporal context information from the CNNs' outputs. The experimental results demonstrate that the transformer-based CNNs can effectively detect COVID-19 on the DiCOVA Track-2 database (AUC: 70.0%) and outperform simple CNNs and hybrid CNN-RNNs.

Classification of Guillain–Barré Syndrome Subtypes Using Sampling Techniques with Binary Approach

Guillain–Barré Syndrome (GBS) is an unusual disorder where the body’s immune system affects the peripheral nervous system. GBS has four main subtypes, whose treatments vary among them. Severe cases of GBS can be fatal. This work aimed to investigate whether balancing an original GBS dataset improves the predictive models created in a previous study. purpleBalancing a dataset is to pursue symmetry in the number of instances of each of the classes.The dataset includes 129 records of Mexican patients diagnosed with some subtype of GBS. We created 10 binary datasets from the original dataset. Then, we balanced these datasets using four different methods to undersample the majority class and one method to oversample the minority class. Finally, we used three classifiers with different approaches to creating predictive models. The results show that balancing the original dataset improves the previous predictive models. The goal of the predictive models is to identify the GBS subtypes applying Machine Learning algorithms. It is expected that specialists may use the model to have a complementary diagnostic using a reduced set of relevant features. Early identification of the subtype will allow starting with the appropriate treatment for patient recovery. This is a contribution to exploring the performance of balancing techniques with real data.

Overlap-Based Undersampling Method for Classification of Imbalanced Medical Datasets

Early diagnosis of some life-threatening diseases such as cancers and heart is crucial for effective treatments. Supervised machine learning has proved to be a very useful tool to serve this purpose. Historical data of patients including clinical and demographic information is used for training learning algorithms. This builds predictive models that provide initial diagnoses. However, in the medical domain, it is common to have the positive class under-represented in a dataset. In such a scenario, a typical learning algorithm tends to be biased towards the negative class, which is the majority class, and misclassify positive cases. This is known as the class imbalance problem. In this paper, a framework for predictive diagnostics of diseases with imbalanced records is presented. To reduce the classification bias, we propose the usage of an overlap-based undersampling method to improve the visibility of minority class samples in the region where the two classes overlap. This is achieved by detecting and removing negative class instances from the overlapping region. This will improve class separability in the data space. Experimental results show achievement of high accuracy in the positive class, which is highly preferable in the medical domain, while good trade-offs between sensitivity and specificity were obtained. Results also show that the method often outperformed other state-of-the-art and well-established techniques.

Machine Learning and Integrative Analysis of Biomedical Big Data

Recent developments in high-throughput technologies have accelerated the accumulation of massive amounts of omics data from multiple sources: genome, epigenome, transcriptome, proteome, metabolome, etc. Traditionally, data from each source (e.g., genome) is analyzed in isolation using statistical and machine learning (ML) methods. Integrative analysis of multi-omics and clinical data is key to new biomedical discoveries and advancements in precision medicine. However, data integration poses new computational challenges as well as exacerbates the ones associated with single-omics studies. Specialized computational approaches are required to effectively and efficiently perform integrative analysis of biomedical data acquired from diverse modalities. In this review, we discuss state-of-the-art ML-based approaches for tackling five specific computational challenges associated with integrative analysis: curse of dimensionality, data heterogeneity, missing data, class imbalance and scalability issues.

An improved support vector machine-based diabetic readmission prediction

BACKGROUND AND OBJECTIVE

In healthcare systems, the cost of unplanned readmission accounts for a large proportion of total hospital payment. Hospital-specific readmission rate becomes a critical issue around the world. Quantification and early identification of unplanned readmission risks will improve the quality of care during hospitalization and reduce the occurrence of readmission. In clinical practice, medical workers generally use LACE score method to evaluate patient readmission risks, but this method usually performs poorly. With this in mind, this study presents a novel method combining support vector machine and genetic algorithm to build the risk prediction model, which simultaneously involves feature selection and the processing of imbalanced data. This model aims to provide decision support for clinicians during the discharge management of patients with diabetes.

METHOD

The experiments were conducted from a set of 8756 medical records with 50 different features about diabetic readmission. After preprocessing the data, an effective SMOTE-based method was proposed to solve the imbalance data problem. Further, in order to improve prediction performance, a hybrid feature selection mechanism was devised to select the important features. Subsequently, an improved support vector machine-based (SVM-based) method was developed and the genetic algorithm was used to tune the sensitive parameter of the algorithm. Finally, the five-fold cross-validation method was applied to compare the performance of proposed method with other methods (LACE score, logistic regression, naïve bayes, decision tree and feed forward neural networks).

RESULTS

Experimental results indicate that the proposed SVM-based method achieves an accuracy of 81.02%, a sensitivity of 82.89%, a specificity of 79.23%, and outperforms other popular algorithms in identifying diabetic patients who may be readmitted.

CONCLUSIONS

Our research can improve the performance of clinic decision support systems for diabetic readmission, by which the readmission possibility as well as the waste of medical resources can be reduced.