Balanced Classification Accuracy Research Articles

Deep Learning-Assisted Discovery of Analogy-Inspired Designs within Peter Collins Analogical Architectural Design Classification Framework

This study focuses on analogical reasoning and deep learning models to enhance the innovative design process in architecture. By constructing multi-layered artificial neural networks, deep learning can derive analogical predictions from structured data to solve complex tasks. Deep learning models interact with analogical thinking patterns in the architectural design process, enabling designers to analyze and draw inspiration from analogical design examples. This study aims to develop a deep learning model that categorizes architectural design examples into specific analogical design classifications. For this purpose, a model based on Convolutional Neural Networks was developed and coded in the Google Colab environment using a dataset of 29,596 visual images, employing Peter Collins' classification system of biological, mechanical, gastronomic, and linguistic analogies. During the training process, the model was trained on images classified according to biological, mechanical, gastronomic, and linguistic categories, achieving an accuracy rate of 98%; however, this rate was recorded as 86% during the testing phase. It was observed that adjustments in the learning rate parameter balanced classification accuracy and training time; lower learning rates reduced accuracy while extending training time. Despite the complexity of architectural images indicated by the 86% accuracy rate on test data, the study emphasizes the model's capacity to achieve accuracy above 95% when confronted with distinct architectural features. In this case, the model allows designers to discover which analogical classification the architectural work to be tested is designed according to, allowing them to develop creative solutions to new design problems. Additionally, this research establishes an interdisciplinary dialogue between artificial intelligence and architecture, providing a foundation for future studies.

Ozone Day Classification using Random Forests with Oversampling and Statistical Tests

Accurate warning of ozone concentration levels in the air is very important for public health. However, the characteristics of the public data related to ozone level detection in the UCI machine learning repository make it difficult to build a warning system based on machine learning techniques. The data consists of 72 relatively large numerical attributes and are measured and collected for 7 years with some blank data, and the distribution of ozone days and normal days is very unbalanced, making it difficult to create an accurate classification model. In this paper to solve the high dimensional attribute problem PCA is applied first, resulting in the 72 attributes being reduced to 20 attributes, and generating slightly better random forests, but the classification for ozone days is still poor due to insufficient data. To solve the insufficient data problem for the minor class which is 6.3% of the total, SMOTE which is one of the representative oversampling methods is applied to a minor class at very high rates repeatedly. It was also checked whether a better machine learning model of random forests can be obtained after applying oversampling at the same very high rate for each class, generating much more synthetic data than the original data and using it to train the random forests. In addition, to ensure the reliability of the synthetic data generated by SMOTE statistical test has been done for each attribute to see if it is statistically reliable. The results of the experiment showed that when the oversampling rate was relatively high with the suggested oversampling and statistical tests, it could be possible to generate synthetic data with statistical characteristics similar to the original data, and by using it to train the random forests, it could be possible to generate random forests with higher and more balanced classification accuracy than using the original data alone, from 94% to 100%. In this sense, this paper has contributed that it provides a methodology to increase the reliability of the machine learning model of random forests for very skewed and high dimensional data like the ozone day classification dataset.

End-to-end deep learning patient level classification of affected territory of ischemic stroke patients in DW-MRI.

To develop an end-to-end DL model for automated classification of affected territory in DWI of stroke patients. In this retrospective multicenter study, brain DWI studies from January 2017 to April 2020 from Center 1, from June 2020 to December 2020 from Center 2, and from November 2019 to April 2020 from Center 3 were included. Four radiologists labeled images into five classes: anterior cerebral artery (ACA), middle cerebral artery (MCA), posterior circulation (PC), and watershed (WS) regions, as well as normal images. Additionally, for Center 1, clinical information was encoded as a domain knowledge vector to incorporate into image embeddings. 3D convolutional neural network (CNN) and attention gate integrated versions for direct 3D encoding, long short-term memory (LSTM-CNN), and time-distributed layer for slice-based encoding were employed. Balanced classification accuracy, macro averaged f1 score, AUC, and interrater Cohen's kappa were calculated. Overall, 624 DWI MRIs from 3 centers were utilized (mean age, interval: 66.89 years, 29-95 years; 345 male) with 439 patients in the training, 103 in the validation, and 82 in the test sets. The best model was a slice-based parallel encoding model with 0.88 balanced accuracy, 0.80 macro-f1 score, and an AUC of 0.98. Clinical domain knowledge integration improved the performance with 0.93 best overall accuracy with parallel stream model embeddings and support vector machine classifiers. The mean kappa value for interrater agreement was 0.87. Developed end-to-end deep learning models performed well in classifying affected regions from stroke in DWI. The end-to-end deep learning model with a parallel stream encoding strategy for classifying stroke regions in DWI has performed comparably with radiologists.

Task-Oriented Over-the-Air Computation for Edge-Device Co-Inference With Balanced Classification Accuracy

Task-Oriented Over-the-Air Computation for Edge-Device Co-Inference With Balanced Classification Accuracy

Predicting histological grade in pediatric glioma using multiparametric radiomics and conventional MRI features

Prediction of glioma is crucial to provide a precise treatment plan to optimize the prognosis of children with glioma. However, studies on the grading of pediatric gliomas using radiomics are limited. Meanwhile, existing methods are mainly based on only radiomics features, ignoring intuitive information about tumor morphology on traditional imaging features. This study aims to utilize multiparametric magnetic resonance imaging (MRI) to identify high-grade and low-grade gliomas in children and establish a classification model based on radiomics features and clinical features. A total of 85 children with gliomas underwent tumor resection, and part of the tumor tissue was examined pathologically. Patients were categorized into high-grade and low-grade groups according to World Health Organization guidelines. Preoperative multiparametric MRI data, including contrast-enhanced T1-weighted imaging, T2-weighted imaging, T2-weighted fluid-attenuated inversion recovery, diffusion-weighted images, and apparent diffusion coefficient sequences, were obtained and labeled by two radiologists. The images were preprocessed, and radiomics features were extracted for each MRI sequence. Feature selection methods were used to select radiomics features, and statistically significant clinical features were identified using t-tests. The selected radiomics features and conventional MRI features were used to train the AutoGluon models. The improved model, based on radiomics features and conventional MRI features, achieved a balanced classification accuracy of 66.59%. The cross-validated areas under the receiver operating characteristic curve for the classifier of AutoGluon frame were 0.8071 on the test dataset. The results indicate that the performance of AutoGluon models can be improved by incorporating conventional MRI features, highlighting the importance of the experience of radiologists in accurately grading pediatric gliomas. This method can help predict the grade of pediatric glioma before pathological examination and assist in determining the appropriate treatment plan, including radiotherapy, chemotherapy, drugs, and gene surgery.

Drop Loss for Person Attribute Recognition With Imbalanced Noisy-Labeled Samples.

Person attribute recognition (PAR) aims to simultaneously predict multiple attributes of a person. Existing deep learning-based PAR methods have achieved impressive performance. Unfortunately, these methods usually ignore the fact that different attributes have an imbalance in the number of noisy-labeled samples in the PAR training datasets, thus leading to suboptimal performance. To address the above problem of imbalanced noisy-labeled samples, we propose a novel and effective loss called drop loss for PAR. In the drop loss, the attributes are treated differently in an easy-to-hard way. In particular, the noisy-labeled candidates, which are identified according to their gradient norms, are dropped with a higher drop rate for the harder attribute. Such a manner adaptively alleviates the adverse effect of imbalanced noisy-labeled samples on model learning. To illustrate the effectiveness of the proposed loss, we train a simple ResNet-50 model based on the drop loss and term it DropNet. Experimental results on two representative PAR tasks (including facial attribute recognition and pedestrian attribute recognition) demonstrate that the proposed DropNet achieves comparable or better performance in terms of both balanced accuracy and classification accuracy over several state-of-the-art PAR methods.

Alzheimer’s disease: using gene/protein network machine learning for molecule discovery in olive oil

Alzheimer’s disease (AD) poses a profound human, social, and economic burden. Previous studies suggest that extra virgin olive oil (EVOO) may be helpful in preventing cognitive decline. Here, we present a network machine learning method for identifying bioactive phytochemicals in EVOO with the highest potential to impact the protein network linked to the development and progression of the AD. A balanced classification accuracy of 70.3 ± 2.6% was achieved in fivefold cross-validation settings for predicting late-stage experimental drugs targeting AD from other clinically approved drugs. The calibrated machine learning algorithm was then used to predict the likelihood of existing drugs and known EVOO phytochemicals to be similar in action to the drugs impacting AD protein networks. These analyses identified the following ten EVOO phytochemicals with the highest likelihood of being active against AD: quercetin, genistein, luteolin, palmitoleate, stearic acid, apigenin, epicatechin, kaempferol, squalene, and daidzein (in the order from the highest to the lowest likelihood). This in silico study presents a framework that brings together artificial intelligence, analytical chemistry, and omics studies to identify unique therapeutic agents. It provides new insights into how EVOO constituents may help treat or prevent AD and potentially provide a basis for consideration in future clinical studies.

Predicting online problem gambling treatment discontinuation: New evidence from cross-validated models.

There are tens of millions of problem gamblers in the world, many of whom either do not seek treatment or fail to commit to it. Dropout rates are high, and not enough is known about factors predicting treatment adherence. We focus on an online cognitive behavioral therapy program for severe problem gambling to determine the likelihood of treatment discontinuation at three different treatment phases: pretreatment, before halfway, and before the end of the program. Participants were Finnish adults (N = 1,139, 670 males, Mage = 34.5) registered in the program between 2019 and 2021. Using logistic regression and five-fold cross-validated naïve Bayes classification, we predicted discontinuation with demographic-, psychometric-, and other gambling-related variables, including the quality of one's social relations, time spent on the waiting list, and experienced readiness to behavioral change. The models had acceptable predictive ability (area under the curve [AUC] values from .69 to .745; cross-validated balanced classification accuracy = 63.2%). In logistic regressions, treatment discontinuation was prominently associated with younger age (p = .008), lower education (p < .001), not being ready to change gambling behavior (p < .001), problem gambling severity (p < .0001), longer time spent on the treatment waiting list (p < .0001), and fewer close social relationships (p < .001). We found significant new real-world evidence on factors statistically predicting treatment discontinuation, which is crucial when existing programs are modified to better serve those in need. (PsycInfo Database Record (c) 2022 APA, all rights reserved).

A novel CO2-EOR potential evaluation method based on BO-LightGBM algorithms using hybrid feature mining

The main application of carbon capture, utilization, and storage (CCUS) in oil-and-gas field development engineering is carbon-dioxide enhanced oil recovery (CO2-EOR). However, accurate and rapid assessment of its application potential remains a major challenge. Traditional methods almost always evaluate whether a reservoir is suitable for the CO2-EOR technology by customizing static screening indicators and establishing threshold ranges. That is irrefutably time consuming, laborious, and subjective. In this study, a novel CO2-EOR potential assessment method based on hybrid feature mining and a light gradient boosting machine using Bayesian optimization (BO-LightGBM) is proposed. The fair cut tree and synthetic minority oversampling technique (FCT-SMOTE) were used to correct the missing and unbalanced oilfield data. This has ensured that the model has a balanced high classification accuracy for each type of reservoir. Hybrid feature selection was used to analyze oilfield data to improve the learning ability of the model and improve the calculation efficiency. To achieve multitasking scalability and rapidity, BO-LightGBM was adopted as a classifier for judging the CO2-EOR potential of reservoirs, and a regressor for predicting EOR factor. The proposed classification model has a prediction accuracy of 98.7%, and the proposed regression model has a root mean square error (RMSE) of 1.72, coefficient of determination (R2) of 0.95, and mean absolute percentage error (MAPE) of 18%. Overall, the CO2-EOR potential assessment method proposed in this study has the advantages of high accuracy, scalability, and speed. In addition, it should be emphasized that using static weight indicators to evaluate CO2-EOR potential may not be appropriate. Rather, machine learning (ML) methods to evaluate the application potential of CO2-EOR is recommended. This is because ML methods can divide the scope of multiple factors simultaneously, providing more intelligent and accurate results.

Hierarchical Multi-Class Classification of Voice Disorders Using Self-Supervised Models and Glottal Features

Previous studies on the automatic classification of voice disorders have mostly investigated the binary classification task, which aims to distinguish pathological voice from healthy voice. Using multi-class classifiers, however, more fine-grained identification of voice disorders can be achieved, which is more helpful for clinical practitioners. Unfortunately, there is little publicly available training data for many voice disorders, which lowers the classification performance on data from unseen speakers. Earlier studies have shown that the usage of glottal source features can reduce data redundancy in detection of laryngeal voice disorders. Another approach to tackle the problems caused by scarcity of training data is to utilize deep learning models, such as wav2vec 2.0 and HuBERT, that have been pre-trained on larger databases. Since the aforementioned approaches have not been thoroughly studied in the multi-class classification of voice disorders, they will be jointly studied in the present work. In addition, we study a hierarchical classifier, which enables task-wise feature optimization and more efficient utilization of data. In this work, the aforementioned three approaches are compared with traditional mel frequency cepstral coefficient (MFCC) features and one-vs-rest and one-vs-one SVM classifiers. The results in a 3-class classification problem between healthy voice and two laryngeal disorders (hyperfunctional dysphonia and vocal fold paresis) indicate that all the studied methods outperform the baselines. The best performance was achieved by using features from wav2vec 2.0 LARGE together with hierarchical classification. The balanced classification accuracy of the system was 62.77% for male speakers, and 55.36% for female speakers, which outperformed the baseline systems by an absolute improvement of 15.76% and 6.95% for male and female speakers, respectively.

Prospective Study of Gray Matter Atrophy Following Pediatric Mild Traumatic Brain Injury.

The clinical and physiologic time course for recovery following pediatric mild traumatic brain injury (pmTBI) remains actively debated. The primary objective of the current study was to prospectively examine structural brain changes (cortical thickness and subcortical volumes) and age-at-injury effects. A priori study hypotheses predicted reduced cortical thickness and hippocampal volumes up to 4 months postinjury, which would be inversely associated with age at injury. Prospective cohort study design with consecutive recruitment. Study inclusion adapted from American Congress of Rehabilitation Medicine (upper threshold) and Zurich Concussion in Sport Group (minimal threshold) and diagnosed by Emergency Department and Urgent Care clinicians. Major neurologic, psychiatric, or developmental disorders were exclusionary. Clinical (Common Data Element) and structural (3 T MRI) evaluations within 11 days (subacute visit [SA]) and at 4 months (early chronic visit [EC]) postinjury. Age- and sex-matched healthy controls (HC) to control for repeat testing/neurodevelopment. Clinical outcomes based on self-report and cognitive testing. Structural images quantified with FreeSurfer (version 7.1.1). A total of 208 patients with pmTBI (age = 14.4 ± 2.9; 40.4% female) and 176 HC (age = 14.2 ± 2.9; 42.0% female) were included in the final analyses (>80% retention). Reduced cortical thickness (right rostral middle frontal gyrus; d = -0.49) and hippocampal volumes (d = -0.24) observed for pmTBI, but not associated with age at injury. Hippocampal volume recovery was mediated by loss of consciousness/posttraumatic amnesia. Significantly greater postconcussive symptoms and cognitive deficits were observed at SA and EC visits, but were not associated with the structural abnormalities. Structural abnormalities slightly improved balanced classification accuracy above and beyond clinical gold standards (∆+3.9%), with a greater increase in specificity (∆+7.5%) relative to sensitivity (∆+0.3%). Current findings indicate that structural brain abnormalities may persist up to 4 months post-pmTBI and are partially mediated by initial markers of injury severity. These results contribute to a growing body of evidence suggesting prolonged physiologic recovery post-pmTBI. In contrast, there was no evidence for age-at-injury effects or physiologic correlates of persistent symptoms in our sample.

A Comparative Study of Deep Learning Algorithms for Detecting Food Intake.

The choice of appropriate machine learning algorithms is crucial for classification problems. This study compares the performance of state-of-the-art time-series deep learning algorithms for classifying food intake using sensor signals. The sensor signals were collected with the help of a wearable sensor system (the Automatic Ingestion Monitor v2, or AIM-2). AIM-2 used an optical and 3-axis accelerometer sensor to capture temporalis muscle activation. Raw signals from those sensors were used to train five classifiers (multilayer perceptron (MLP), time Convolutional Neural Network (time-CNN), Fully Convolutional Neural Network (FCN), Residual Neural Network (ResNet), and Inception network) to differentiate food intake (eating and drinking) from other activities. Data were collected from 17 pilot subjects over the course of 23 days in free-living conditions. A leave one subject out cross-validation scheme was used for training and testing. Time-CNN, FCN, ResNet, and Inception achieved average balanced classification accuracy of 88.84%, 90.18%, 93.47%, and 92.15%, respectively. The results indicate that ResNet outperforms other state-of-the-art deep learning algorithms for this specific problem.

Classifying Kepler light curves for 12 000 A and F stars using supervised feature-based machine learning

ABSTRACT With the availability of large-scale surveys like Kepler and TESS, there is a pressing need for automated methods to classify light curves according to known classes of variable stars. We introduce a new algorithm for classifying light curves that compares 7000 time-series features to find those that most effectively classify a given set of light curves. We apply our method to Kepler light curves for stars with effective temperatures in the range 6500–10 000 K. We show that the sample can be meaningfully represented in an interpretable 5D feature space that separates seven major classes of light curves (δ Scuti stars, γ Doradus stars, RR Lyrae stars, rotational variables, contact eclipsing binaries, detached eclipsing binaries, and non-variables). We achieve a balanced classification accuracy of 82 per cent on an independent test set of Kepler stars using a Gaussian mixture model classifier. We use our method to classify 12 000 Kepler light curves from Quarter 9 and provide a catalogue of the results. We further outline a confidence heuristic based on probability density to search our catalogue and extract candidate lists of correctly classified variable stars.

Metabolite selection for machine learning in childhood brain tumour classification.

MRS can provide high accuracy in the diagnosis of childhood brain tumours when combined with machine learning. A feature selection method such as principal component analysis is commonly used to reduce the dimensionality of metabolite profiles prior to classification. However, an alternative approach of identifying the optimal set of metabolites has not been fully evaluated, possibly due to the challenges of defining this for a multi-class problem. This study aims to investigate metabolite selection from in vivo MRS for childhood brain tumour classification. Multi-site 1.5T and 3T cohorts of patients with a brain tumour and histological diagnosis of ependymoma, medulloblastoma and pilocytic astrocytoma were retrospectively evaluated. Dimensionality reduction was undertaken by selecting metabolite concentrations through multi-class receiver operating characteristics and compared with principal component analysis. Classification accuracy was determined through leave-one-out and k-fold cross-validation. Metabolites identified as crucial in tumour classification include myo-inositol (P < 0.05, ), total lipids and macromolecules at 0.9ppm (P < 0.05, ) and total creatine (P < 0.05, ) for the 1.5T cohort, and glycine (P < 0.05, ), total N-acetylaspartate (P < 0.05, ) and total choline (P < 0.05, ) for the 3T cohort. Compared with the principal components, the selected metabolites were able to provide significantly improved discrimination between the tumours through most classifiers (P < 0.05). The highest balanced classification accuracy determined through leave-one-out cross-validation was 85% for 1.5T 1 H-MRS through support vector machine and 75% for 3T 1 H-MRS through linear discriminant analysis after oversampling the minority. The study suggests that a group of crucial metabolites helps to achieve better discrimination between childhood brain tumours.

Transcriptomics and machine learning to advance schizophrenia genetics: A case-control study using post-mortem brain data

Background and ObjectiveAlterations of the expression of a variety of genes have been reported in patients with schizophrenia (SCZ). Moreover, machine learning (ML) analysis of gene expression microarray data has shown promising preliminary results in the study of SCZ. Our objective was to evaluate the performance of ML in classifying SCZ cases and controls based on gene expression microarray data from the dorsolateral prefrontal cortex. MethodsWe apply a state-of-the-art ML algorithm (XGBoost) to train and evaluate a classification model using 201 SCZ cases and 278 controls. We utilized 10-fold cross-validation for model selection, and a held-out testing set to evaluate the model. The performance metric utilizes to evaluate classification performance was the area under the receiver-operator characteristics curve (AUC). ResultsWe report an average AUC on 10-fold cross-validation of 0.76 and an AUC of 0.76 on testing data, not used during training. Analysis of the rolling balanced classification accuracy from high to low prediction confidence levels showed that the most certain subset of predictions ranged between 80-90%. The ML model utilized 182 gene expression probes. Further improvement to classification performance was observed when applying an automated ML strategy on the 182 features, which achieved an AUC of 0.79 on the same testing data. We found literature evidence linking all of the top ten ML ranked genes to SCZ. Furthermore, we leveraged information from the full set of microarray gene expressions available via univariate differential gene expression analysis. We then prioritized differentially expressed gene sets using the piano gene set analysis package. We augmented the ranking of the prioritized gene sets with genes from the complex multivariate ML model using hypergeometric tests to identify more robust gene sets. We identified two significant Gene Ontology molecular function gene sets: “oxidoreductase activity, acting on the CH-NH2 group of donors” and “integrin binding.” Lastly, we present candidate treatments for SCZ based on findings from our study ConclusionsOverall, we observed above-chance performance from ML classification of SCZ cases and controls based on brain gene expression microarray data, and found that ML analysis of gene expressions could further our understanding of the pathophysiology of SCZ and help identify novel treatments.

Meta-Analysis of Apparent Diffusion Coefficient in Pediatric Medulloblastoma, Ependymoma, and Pilocytic Astrocytoma.

Medulloblastoma, ependymoma, and pilocytic astrocytoma are common pediatric posterior fossa tumors. These tumors show overlapping characteristics on conventional MRI scans, making diagnosis difficult. To investigate whether apparent diffusion coefficient (ADC) values differ between tumor types and to identify optimum cut-off values to accurately classify the tumors using different performance metrics. Systematic review and meta-analysis. Seven studies reporting ADC in pediatric posterior fossa tumors (115 medulloblastoma, 68 ependymoma, and 86 pilocytic astrocytoma) were included following PubMed and ScienceDirect searches. Diffusion weighted imaging (DWI) was performed on 1.5 and 3T across multiple institution and vendors. The combined mean and standard deviation of ADC were calculated for each tumor type using a random-effects model, and the effect size was calculated using Hedge's g. Sensitivity/specificity, weighted classification accuracy, balanced classification accuracy. A P value < 0.05 was considered statistically significant, and a Hedge's g value of >1.2 was considered to represent a large difference. The mean (± standard deviation) ADCs of medulloblastoma, ependymoma, and pilocytic astrocytoma were 0.76 ± 0.16, 1.10 ± 0.10, and 1.49 ± 0.16 mm2 /sec× 10-3 . To maximize sensitivity and specificity using the mean ADC, the cut-off was found to be 0.96 mm2 /sec× 10-3 for medulloblastoma and ependymoma and 1.26 mm2 /sec× 10-3 for ependymoma and pilocytic astrocytoma. The meta-analysis showed significantly different ADC distributions for the three posterior fossa tumors. The cut-off values changed markedly (up to 7%) based on the performance metric used and the prevalence of the tumor types. There were significant differences in ADC between tumor types. However, it should be noted that only summary statistics from each study were analyzed and there were differences in how regions of interest were defined between studies. 1 TECHNICAL EFFICACY: Stage 3.

SPECT Image Features for Early Detection of Parkinson's Disease using Machine Learning Methods.

Millions of people around the world suffer from Parkinson's disease, a neurodegenerative disorder with no remedy. Currently, the best response to interventions is achieved when the disease is diagnosed at an early stage. Supervised machine learning models are a common approach to assist early diagnosis from clinical data, but their performance is highly dependent on available example data and selected input features. In this study, we explore 23 single photon emission computed tomography (SPECT) image features for the early diagnosis of Parkinson's disease on 646 subjects. We achieve 94 % balanced classification accuracy in independent test data using the full feature space and show that matching accuracy can be achieved with only eight features, including original features introduced in this study. All the presented features can be generated using a routinely available clinical software and are therefore straightforward to extract and apply.

A three-branch 3D convolutional neural network for EEG-based different hand movement stages classification

Motor Imagery is a classical method of Brain Computer Interaction, in which electroencephalogram (EEG) signal features evoked by the imaginary body movements are recognized, and relevant information is extracted. Recently, various deep learning methods are being focused on finding an easy-to-use EEG representation method that can preserve both temporal information as well as spatial information. To further utilize the spatial and temporal features of EEG signals, we proposed a 3D representation of EEG and an end-to-end EEG three-branch 3D convolutional neural network, to address the class imbalance problem (dataset show unequal distribution among their classes), we proposed a class balance cropped strategy. Experimental results indicated that there are also a problem of the different classification difficulty for different classes in motor stages classification tasks, we introduce focal loss to address problem of ‘easy-hard’ examples, when trained with the focal loss, the three-branch 3D-CNN network achieve good performance (relatively more balanced classification accuracy of binary classifications) on the WAY-EEG-GAL data set. Experimental results show that the proposed method is a good method, which can improve classification effect of different motor stages classification.

Alzheimer's classification using dynamic ensemble of classifiers selection algorithms: A performance analysis

Alzheimer's is a type of severe cognitive impairment where an individual cannot do their daily day-to-day activities. It is a challenging task to find out the Alzheimer's and Mild Cognitive Impairment patients. This study aims to compare the performance of the state of the art Dynamic Ensemble Selection of Classifier algorithms for classifying healthy, Mild Cognitive Impairment, and Alzheimer's disease participants at the baseline stage itself using multimodal features. The data used in the study is from Alzheimer's Disease Neuroimaging Initiative-TADPOLE dataset. The medical imaging, Cerebro-spinal fluid, cognitive test, and demographics data of the patients at the baseline visits are considered for the prediction purpose. The performance of the state-of-the-art Dynamic Ensemble of Classifier Selection algorithms is compared using these features in terms of Balanced Classification Accuracy, Sensitivity, and Specificity. The most commonly used pool of Machine Learning classifiers is used as the input for Dynamic Ensemble of Classifier Selection algorithms. Moreover, the performance of the pool of Machine Learning classifiers without using the Dynamic Ensemble Selection of Classifiers algorithms are also compared. The performance metrics such as Balanced Classification Accuracy, Sensitivity, and Specificity are increased after using the Dynamic Ensemble of Classifier Selection algorithms on most of the pool of classifiers for classifying healthy, Alzheimer's, and Mild Cognitive Impairment patients is promising.

Reconstituting T Cell Receptor Selection In-Silico

Abstract Each T cell receptor (TCR) gene is created without regard for which substances (antigens) the receptor can recognize. T cell selection culls developing T cells when their TCRs (i) fail to recognize major histocompatibility complexes (MHCs) that act as antigen presenting platforms or (ii) recognize with high affinity self-antigens derived from healthy cells and tissue. Here, we show that we can accurately predict from the TCR if any T cell is pre- or post-selection (and by extension if that T cell would be culled). First, we analyze millions of TCR genes sequenced from mouse spleen using high-throughput receptor sequencing. To get examples of TCRs pre-selection, we develop a way to translate non-productive TCR genes that otherwise cannot express protein, obtaining candidate TCR protein sequences never subjected to T cell selection. To get examples of TCRs post-selection, we translate productive TCR genes sequenced from a peripheral location such as spleen, obtaining TCR protein sequences representing T cells that survived T cell selection. Next, we train a machine learning model to find sequence level patterns that discriminate between these two categories of TCR protein sequences, achieving a balanced classification accuracy of 73% (AUC of 0.8) on de-novo TCR β-chain protein sequences. This classification accuracy may be near the achievable limit using only the β-chain representing half of the TCR heterodimer. To verify our approach, we show our machine learning model accurately predicts that (i) productive TCRs from thymus (where T selection occurs) are predominantly pre-selection and (ii) productive TCRs from peripheral tissues like skin are post-selection.