Binary Classification Problem Research Articles

Interpretable machine learning for predicting urban flash flood hotspots using intertwined land and built-environment features

Pluvial flash floods are fast-moving hazards and causes significant disruptions in urban areas. With the increase in heavy precipitations, the ability to proactively identify flash floods hotspots in cities is critical for flood nowcasting and predictive monitoring of risks. While rainfall runoff models and hydrologic models are useful models for flash flood prediction, these models are computationally expensive and effort intensive to be used for flood nowcasting. To address this challenge, this study presents interpretable machine learning models for predicting urban flash flood hotspots based on intertwined land and built environment features. The task of predicting flash flood hotspots is formulated as a binary classification problem, and three recent flash flood events in U.S. cities are selected for data collection and model validation. Various features related to land and built environment characteristics are constructed using diverse datasets, and the occurrences of flash floods are captured using crowdsource data from the events. Using these features and datasets, the flash flood hotspots of cities are predicted with two ensemble models based on decision trees. The results demonstrate that the models can achieve good accuracy (0.8) in identifying flooded/non-flooded locations. Especially, the models can achieve high true positive rate (0.83–0.89) and low missing rate, demonstrating the methods' practicability for accurately predicting flooded hotspots. The model interpretation results indicate that land features related to hydrological and topological features have greater impacts on flash flood risk, than built environment features. Further analysis reveals that the feature importance, model performance, and model transferability performance vary among cities and localized specifications of the models are needed for accurate prediction of flash flood for a particular city. The data-driven machine learning models presented in this study provide a useful tool for predicting flash flood hotspots based on the intertwined features of land and the built environment in cities to enable nowcasting and proactive monitoring of flash flood hotspots for emergency response and also inform integrated urban design and development towards flash flood risk reduction.

Opportunities to Reduce the Risk of Cardiovascular Death by Improving Machine Learning Methods

Improving algorithms for automatic recognition of electrocardiograms requires increasing of training dataset, which is not always possible due to the rarity of certain cardiac diseases or ethical issues. It is possible to improve the algorithms for generating synthetic electrocardiograms using mechanistic models and generative-descriptive neural networks (GANs). At the same time, when evaluating the effectiveness of the proposed solutions, various authors offer different quality assessment metrics from subjective expert assessment to the squared mean error. We compare two approaches to generating synthetic electrocardiograms: pseudo-ECG generation using a one-dimensional cardiomyocyte model and GAN based on long-term short-term memory in terms of machine learning metrics: accuracy, recall, f1-score. We solve the problem of binary classification with bagging method, class 0 — normal sinus rhythm, class 1 — atrial fibrillation. We found that classifier trained on synthetic ECGs generated using GAN is slightly more effective compared to pseudo-ECGs generated using mechanistic models ones. At the same time, it turned out that GANs are not stable enough. The generation of synthetic electrocardiograms using both mechanistic models and GAN can be used to enrich the training set in case of recognition of rare cardiac diseases.

Superpixel-wise contrast exploration for salient object detection

Salient object detection (SOD) methods typically consider SOD as a pixel-wise binary classification problem and utilize the binary cross-entropy (BCE) loss for optimization. However, the BCE loss ignores the global dependencies between pixels of the entire image, which is important for ensuring the accuracy and integrity of objects. To address this limitation explicitly, contrastive learning is introduced to enhance both the intra-pixel compactness of the foreground and inter-pixel separability between the foreground and background. Unfortunately, the random pixel sampling strategies of existing contrast algorithms cannot ensure that the sampled pixels cover all semantic contents across the entire image because of the lack of fine-grained semantic categories for pixels in SOD. Hence, in this study, a SuperPixel-wise Contrastive (SPCont) algorithm is proposed, which samples all superpixel centers to improve the diversity and representativeness of the sampled pixels. In addition, a SOD-specific hard pixel sampling strategy is designed to refine the edge details of the objects. An intra-image multi-level negative sampling approach is presented to increase the number of negative samples. Extensive experiments are conducted to demonstrate that incorporating the SPCont algorithm into the state-of-the-art network results in significant performance improvements. We expect that our work will open new paths for the application of contrastive learning to SOD.

Artificial intelligence in tongue diagnosis: classification of tongue lesions and normal tongue images using deep convolutional neural network

ObjectiveThis study aims to classify tongue lesion types using tongue images utilizing Deep Convolutional Neural Networks (DCNNs).MethodsA dataset consisting of five classes, four tongue lesion classes (coated, geographical, fissured tongue, and median rhomboid glossitis), and one healthy/normal tongue class, was constructed using tongue images of 623 patients who were admitted to our clinic. Classification performance was evaluated on VGG19, ResNet50, ResNet101, and GoogLeNet networks using fusion based majority voting (FBMV) approach for the first time in the literature.ResultsIn the binary classification problem (normal vs. tongue lesion), the highest classification accuracy performance of 93,53% was achieved utilizing ResNet101, and this rate was increased to 95,15% with the application of the FBMV approach. In the five-class classification problem of tongue lesion types, the VGG19 network yielded the best accuracy rate of 83.93%, and the fusion approach improved this rate to 88.76%.ConclusionThe obtained test results showed that tongue lesions could be identified with a high accuracy by applying DCNNs. Further improvement of these results has the potential for the use of the proposed method in clinic applications.

Comparison and analysis of the effect of various machine learning algorithms on cancer prediction

Cancer is a serious disease that can have a big impact on the physical and mental health of patients. The incidence and mortality rate of cancer are increasing globally. Therefore, predicting the occurrence and treatment effect of cancer has become a hot research topic in the medical field. Machine learning algorithms can use vast amounts of data and algorithmic models to predict aspects of cancer occurrence, progression, and treatment effectiveness. This algorithm can learn and find patterns from large amounts of medical data, thereby improving the diagnosis and treatment of cancer. In this paper, naive Bayes model, logistic regression model, random forest model, KNN model and Ridge regression model were used to predict the data of cancer patients. The accuracy of naive Bayes model and logistic regression model is the highest, reaching 95%. The accuracy of random forest model and KNN model is 94%. The worst performing model was Ridge regression, with an accuracy of 93%. Naive Bayes algorithm and logistic regression algorithm perform well in solving binary classification problems because they are both probability-based classification algorithms. In this problem, the number of features is large, so the ridge regression algorithm can effectively process the data, but due to the small amount of data, it is easy to appear underfitting, so the performance is poor. In this paper, a variety of machine learning algorithms are used to predict cancer patients, and the accuracy, accuracy, recall rate and f1 parameters of each model are calculated and compared, which provides a basis for subsequent research.

Application of Machine Learning in the Quantitative Analysis of the Surface Characteristics of Highly Abundant Cytoplasmic Proteins: Toward AI-Based Biomimetics.

Proteins in the crowded environment of human cells have often been studied regarding nonspecific interactions, misfolding, and aggregation, which may cause cellular malfunction and disease. Specifically, proteins with high abundance are more susceptible to these issues due to the law of mass action. Therefore, the surfaces of highly abundant cytoplasmic (HAC) proteins directly exposed to the environment can exhibit specific physicochemical, structural, and geometrical characteristics that reduce nonspecific interactions and adapt to the environment. However, the quantitative relationships between the overall surface descriptors still need clarification. Here, we used machine learning to identify HAC proteins using hydrophobicity, charge, roughness, secondary structures, and B-factor from the protein surfaces and quantified the contribution of each descriptor. First, several supervised learning algorithms were compared to solve binary classification problems for the surfaces of HAC and extracellular proteins. Then, logistic regression was used for the feature importance analysis of descriptors considering model performance (80.2% accuracy and 87.6% AUC) and interpretability. The HAC proteins showed positive correlations with negatively and positively charged areas but negative correlations with hydrophobicity, the B-factor, the proportion of beta structures, roughness, and the proportion of disordered regions. Finally, the details of each descriptor could be explained concerning adaptative surface strategies of HAC proteins to regulate nonspecific interactions, protein folding, flexibility, stability, and adsorption. This study presented a novel approach using various surface descriptors to identify HAC proteins and provided quantitative design rules for the surfaces well-suited to human cellular crowded environments.

Ultra-high-resolution UAV-imaging and supervised deep learning for accurate detection of Alternaria solani in potato fields.

Alternaria solani is the second most devastating foliar pathogen of potato crops worldwide, causing premature defoliation of the plants. This disease is currently prevented through the regular application of detrimental crop protection products and is guided by early warnings based on weather predictions and visual observations by farmers. To reduce the use of crop protection products, without additional production losses, it would be beneficial to be able to automatically detect Alternaria solani in potato fields. In recent years, the potential of deep learning in precision agriculture is receiving increasing research attention. Convolutional Neural Networks (CNNs) are currently the state of the art, but also come with challenges, especially regarding in-field robustness. This stems from the fact that they are often trained on datasets that are limited in size or have been recorded in controlled environments, not necessarily representative of real-world settings. We collected a dataset consisting of ultra-high-resolution modified RGB UAV-imagery of both symptomatic and non-symptomatic potato crops in the field during various years and disease stages to cover the great variability in agricultural data. We developed a convolutional neural network to perform in-field detection of Alternaria, defined as a binary classification problem. Our model achieves a similar accuracy as several state-of-the-art models for disease detection, but has a much lower inference time, which enhances its practical applicability. By using training data of three consecutive growing seasons (2019, 2020 and 2021) and test data of an independent fourth year (2022), an F1 score of 0.93 is achieved. Furthermore, we evaluate how different properties of the dataset such as its size and class imbalance impact the obtained accuracy.

Optimizer algorithms and convolutional neural networks for text classification

<span lang="EN-US">Lately, deep learning has improved the algorithms and the architectures of several natural language processing (NLP) tasks. In spite of that, the performance of any deep learning model is widely impacted by the used optimizer algorithm; which allows updating the model parameters, finding the optimal weights, and minimizing the value of the loss function. Thus, this paper proposes a new convolutional neural network (CNN) architecture for text classification (TC) and sentiment analysis and uses it with various optimizer algorithms in the literature. Actually, in NLP, and particularly for sentiment classification concerns, the need for more empirical experiments increases the probability of selecting the pertinent optimizer. Hence, we have evaluated various optimizers on three types of text review datasets: small, medium, and large. Thereby, we examined the optimizers regarding the data amount and we have implemented our CNN model on three different sentiment analysis datasets so as to binary label text reviews. The experimental results illustrate that the adaptive optimization algorithms Adam and root mean square propagation (RMSprop) have surpassed the other optimizers. Moreover, our best CNN model which employed the RMSprop optimizer has achieved 90.48% accuracy and surpassed the state-of-the-art CNN models for binary sentiment classification problems.</span>

DEVELOPMENT OF A BINARY CLASSIFICATION MODEL BASED ON SMALL DATA USING MACHINE LEARNING METHODS

Today, solutions to the problem of binary classification using machine learning find applications in a huge number of spheres of life, such as medicine, energy, marketing, agriculture, financial analytics, etc. This is a great opportunity for companies to gain new sources of profit and improve existing processes. Therefore, new solution methods are being actively developed, existing ones are being improved, and research is being conducted on the possibility of using machine learning to solve classification problems in various fields. The study of the effectiveness of using various machine learning methods, taking into account the existing problems of small data in solving the problem of binary classification, is very relevant due to the significant preponderance of developments towards Big Data. For small data, possible problems that affect the effectiveness of the trained model have been identified, and various options for solving these problems have been proposed. To assess the impact of small data problems on the quality of the trained model, a comparative analysis of the quality metrics of models trained on different variations of data processing was carried out. It is concluded that correct work with small data requires timely elimination of such data defects as class imbalance, outliers, etc. In the course of the study, the most significant quality metrics were selected to obtain a model for analyzing medical parameters. A comparative analysis of diabetes detection models based on preprocessed small data has been carried out. For the task under consideration, the stacking model was chosen as the best option for medical use. The results of the analysis showed that machine learning is able to show high efficiency in solving real problems of binary classification.

Application of the AlphaFold2 Protein Prediction Algorithm Based on Artificial Intelligence

As the expression products of genes and macromolecules in living organisms, proteins are the main material basis of life activities. They exist widely in various cells and have various functions such as catalysis, cell signaling and structural support, playing a key role in life activities and functional execution. At the same time, the study of protein can better grasp the life activities from the molecular level, and has important practical significance for disease management, new drug development and crop improvement. Due to advances in high-throughput sequencing technology, protein sequence data has grown exponentially. The protein function prediction problem can be seen as a multi-label binary classification problem by extracting the features of a given protein and mapping them to the protein function label space. A variety of data sources can be mined to obtain protein function prediction features, such as protein sequence, protein structure, protein family, protein interaction network, etc. The initial steps are classical sequence-based methods, such as BLAST, which calculate the similarity between protein sequences and transmit annotations between proteins whose similarity scores exceed a specific threshold. This method has great limitations for protein function prediction without sequence similarity. Therefore, this paper analyzes the development prospect of bioanalysis and artificial intelligence through the application status and realization path of AlphaFold2 protein prediction algorithm based on artificial intelligence.

A distance-based kernel for classification via Support Vector Machines.

Support Vector Machines (SVMs) are a type of supervised machine learning algorithm widely used for classification tasks. In contrast to traditional methods that split the data into separate training and testing sets, here we propose an innovative approach where subsets of the original data are randomly selected to train the model multiple times. This iterative training process aims to identify a representative data subset, leading to improved inferences about the population. Additionally, we introduce a novel distance-based kernel specifically designed for binary-type features based on a similarity matrix that efficiently handles both binary and multi-class classification problems. Computational experiments on publicly available datasets of varying sizes demonstrate that our proposed method significantly outperforms existing approaches in terms of classification accuracy. Furthermore, the distance-based kernel achieves superior performance compared to other well-known kernels from the literature and those used in previous studies on the same datasets. These findings validate the effectiveness of our proposed classification method and distance-based kernel for SVMs. By leveraging random subset selection and a unique kernel design, we achieve notable improvements in classification accuracy. These results have significant implications for diverse classification problems in Machine Learning and data analysis.

Automatic evaluation of the Nine-Hole Peg Test in multiple sclerosis patients using machine learning models

Neurological damage in Multiple Sclerosis (MS) affects the motor, sensory, visual, cognitive and sphincteric systems, resulting in different degrees of disability. Upper limb dysfunction is a core deficit affecting MS patients. However, since upper limb function impairment is not properly recorded in the Expanded Disability Status Scale (EDSS) for MS, other tests have been developed in the last decades. The Nine Hole Peg Test (9HPT) is widely considered a gold standard metric for manual dexterity. Unfortunately, its quantification strongly depends on the subjectivity of each clinician.The aim of this work is to standardize the measurement process of the 9HPT. Instead of a human interpretation of the time needed to complete the exercise, this paper aims to objectively extract multiple kinematic information of the subject’s hand (period, velocity, time). This information is used to detect the presence and the state of the disease more accurately and classify the patients automatically. The manuscript describes a developed framework to automate and objectivize the 9HPT. The framework evaluates the performance of different Machine Learning models and Artificial Neural Networks on MS diagnosis. The built system analyzes videos of subjects performing the 9HPT, extracts the hand’s pose using Convolutional Neural Networks, acquires the hand’s kinematic features and classifies the severity of the disease using Machine Learning algorithms. For the framework validation, two classification approaches are proposed. The first one is a binary classification problem, separating the subjects between control and patients. The second one is a multiclass classification problem, where the patients are separated into three different classes based on the severity assessed by the clinicians. A technical validation has been implemented with an accuracy of up 96.67% for the binary classification and 92.22% for the multiclass classification.

On Optimizing Ensemble Models using Column Generation

AbstractIn recent years, an interest appeared in integrating various optimization algorithms in machine learning. We study the potential of ensemble learning in classification tasks and how to efficiently decompose the underlying optimization problem. Ensemble learning has become popular for machine learning applications and it is particularly interesting from an optimization perspective due to its resemblance to column generation. The challenge for learning is not only to obtain a good fit for the training data set, but also good generalization, such that the classifier is generally applicable. Deep networks have the drawback that they require a lot of computational effort to get to an accurate classification. Ensemble learning can combine various weak learners, which individually require less computational time. We consider binary classification problems studying a three-phase algorithm. After initializing a set of base learners refined by a bootstrapping approach, base learners are generated using the solution of an linear programming (LP) master problem and then solving a machine learning sub-problem regarding a reduced data set, which can be viewed as a so-called pricing problem. We theoretically show that the algorithm computes an optimal ensemble model in the convex hull of a given model space. The implementation of the algorithm is part of an ensemble learning framework called decolearn. Numerical experiments with CIFAR-10 data set show that the base learners are diverse and that both the training and generalization error are reduced after several iterations.

Analysis of Emotional Tendency Based on Chinese Sugar-water Shop Evaluation Text

The catering market has developed rapidly, and under the influence of the Internet and the epidemic, online consumption has become increasingly strong. Food is the life of the people, and sweets are an important factor in improving happiness in life. For the sugar water store, combined with online and offline fine operation, grasp the user evaluation, in order to stand out in the competition. This research uses the evaluation data of the eight sugar water shops in Guangzhou to extract keywords using TextRank for the texts with long evaluation data, so that the evaluation texts are controlled within 200 words, and then machine learning algorithms are used to analyze, mine and classify them. The experimental results show that this method can solve the binary classification problem of positive emotion and negative emotion in short time.

Credit and Loan Approval Classification Using a Bio-Inspired Neural Network.

Numerous people are applying for bank loans as a result of the banking industry's expansion, but because banks only have a certain amount of assets to lend to, they can only do so to a certain number of applicants. Therefore, the banking industry is very interested in finding ways to reduce the risk factor involved in choosing the safe applicant in order to save lots of bank resources. These days, machine learning greatly reduces the amount of work needed to choose the safe applicant. Taking this into account, a novel weights and structure determination (WASD) neural network has been built to meet the aforementioned two challenges of credit approval and loan approval, as well as to handle the unique characteristics of each. Motivated by the observation that WASD neural networks outperform conventional back-propagation neural networks in terms of sluggish training speed and being stuck in local minima, we created a bio-inspired WASD algorithm for binary classification problems (BWASD) for best adapting to the credit or loan approval model by utilizing the metaheuristic beetle antennae search (BAS) algorithm to improve the learning procedure of the WASD algorithm. Theoretical and experimental study demonstrate superior performance and problem adaptability. Furthermore, we provide a complete MATLAB package to support our experiments together with full implementation and extensive installation instructions.

GRRLN: Gated Recurrent Residual Learning Networks for code clone detection

AbstractCode clone detection is a critical problem in software development and maintenance domains. It aims to identify functionally identical or similar code fragments within an application. Existing works formulate the code clone detection task as a binary classification problem which predicts a code pair as a clone or not based on a pre‐defined threshold. In reality, there are various types of code clone subject to the degree of how a pair of code fragments are similar to each other. To investigate the effect of different code clone detection manners on the clone detection result, we propose Gated Recurrent Residual Learning Networks (GRRLN), a novel neural network model for code clone detection. To train GRRLN, we first represent each code fragment as a statement‐level tree sequence derived from the whole abstract syntax tree (AST). Then, a gated recurrent neural network with residual connections is adopted to fully extract the semantics of all individual statement trees together with their dependency relationships across the input statement sequence. Finally, the output representations of code fragments by GRRLN are used for similarity calculation and clone detection. We evaluate GRRLN using two real‐world datasets for code clone detection and clone type classification. Experiments show that GRRLN achieves promising and compelling results and meanwhile needs significantly less time and memory consumption compared with the state‐of‐the‐art methods.

Ensemble methods of rank-based trees for single sample classification with gene expression profiles

Building Single Sample Predictors (SSPs) from gene expression profiles presents challenges, notably due to the lack of calibration across diverse gene expression measurement technologies. However, recent research indicates the viability of classifying phenotypes based on the order of expression of multiple genes. Existing SSP methods often rely on Top Scoring Pairs (TSP), which are platform-independent and easy to interpret through the concept of “relative expression reversals”. Nevertheless, TSP methods face limitations in classifying complex patterns involving comparisons of more than two gene expressions. To overcome these constraints, we introduce a novel approach that extends TSP rules by constructing rank-based trees capable of encompassing extensive gene-gene comparisons. This method is bolstered by incorporating two ensemble strategies, boosting and random forest, to mitigate the risk of overfitting. Our implementation of ensemble rank-based trees employs boosting with LogitBoost cost and random forests, addressing both binary and multi-class classification problems. In a comparative analysis across 12 cancer gene expression datasets, our proposed methods demonstrate superior performance over both the k-TSP classifier and nearest template prediction methods. We have further refined our approach to facilitate variable selection and the generation of clear, precise decision rules from rank-based trees, enhancing interpretability. The cumulative evidence from our research underscores the significant potential of ensemble rank-based trees in advancing disease classification via gene expression data, offering a robust, interpretable, and scalable solution. Our software is available at https://CRAN.R-project.org/package=ranktreeEnsemble.

Two‐stage credit risk prediction framework based on three‐way decisions with automatic threshold learning

AbstractCredit risk prediction is a binary classification problem. Using two‐way decisions to classify defaulters may lead to decision errors due to insufficient information. To solve this issue, in addition to identifying borrowers as defaulters and nondefaulters, this paper introduced the delay‐decision mechanism in three‐way decisions, so that records acquiring more information do not make decisions immediately. A two‐stage credit risk prediction framework based on three‐way decisions was proposed to reduce decision risk. In this framework, the decision cost values of three‐way decisions were simplified by analyzing the credit risk prediction, and the expression of threshold calculation was also modified. An optimization objective was built according to the trade‐off between information gain and decision cost, and the particle swarm optimization algorithm was applied to learn the decision thresholds. After adding more supplementary information, the samples in the delayed‐decision region were made further decisions. A dataset from a commercial bank in China was employed to conduct experiments, and the results demonstrated that our proposed method outperformed various base classifiers.

Binarizing by Classification: Is Soft Function Really Necessary?

Binary neural networks leverage Sign function to binarize weights and activations, which require gradient estimators to overcome its non-differentiability and will inevitably bring gradient errors during backpropagation. Although many hand-designed soft functions have been proposed as gradient estimators to better approximate gradients, their mechanism is not clear and there are still huge performance gaps between binary models and their full-precision counterparts. To address these issues and reduce gradient error, we propose to tackle network binarization as a binary classification problem and use a multi-layer perceptron (MLP) as the classifier in the forward pass and gradient estimator in the backward pass. Benefiting from the MLP’s theoretical capability to fit any continuous function, it can be adaptively learned to binarize networks and backpropagate gradients without any prior knowledge of soft functions. From this perspective, we further empirically justify that even a simple linear function can outperform previous complex soft functions. Extensive experiments demonstrate that the proposed method yields surprising performance both in image classification and human pose estimation tasks. Specifically, we achieve 65.7% top-1 accuracy of ResNet-34 on ImageNet dataset, with an absolute improvement of 2.6%. Moreover, we take binarization as a lightweighting approach for pose estimation models and propose well-designed binary pose estimation networks SBPN and BHRNet. When evaluating on the challenging Microsoft COCO keypoint dataset, the proposed method enables binary networks to achieve a mAP of up to 60.6 for the first time. Experiments conducted on real platforms demonstrate that BNN achieves a better balance between performance and computational complexity, especially when computational resources are extremely low.

Deep learning-based predictive models for forex market trends: Practical implementation and performance evaluation.

In recent years, there has been growing interest in the prediction of financial market trends, due to its potential applications in the real world. Unlike traditional investment avenues such as the stock market, the foreign exchange (Forex) market revolves around two primary types of orders that correspond with the market's direction: upward and downward. Consequently, forecasting the behaviour of the Forex behaviour market can be simplified into a binary classification problem to streamline its complexity. Despite the significant enhancements and improvements in performance seen in recent proposed predictive models for the forex market, driven by the advancement of deep learning in various domains, it remains imperative to approach these models with careful consideration of best practices and real-world applications. Currently, only a limited number of papers have been dedicated to this area. This article aims to bridge this gap by proposing a practical implementation of deep learning-based predictive models that perform well for real-world trading activities. These predictive mechanisms can help traders in minimising budget losses and anticipate future risks. Furthermore, the paper emphasises the importance of focussing on return profit as the evaluation metric, rather than accuracy. Extensive experimental studies conducted on realistic Yahoo Finance data sets validate the effectiveness of our implemented prediction mechanisms. Furthermore, empirical evidence suggests that employing the use of three-value labels yields superior accuracy performance compared to traditional two-value labels, as it helps reduce the number of orders placed.