Classifier Training Research Articles

A facial recognition using a combination of a novel one dimension deep CNN and LDA

Face recognition is one of the useful tasks and can be used for many applications as security systems, it is necessary to find effective and low complexity facial classifier methods. In this paper, we proposed a new one-dimensional CNN deep convolutional neural network (1D-DCNN) classifier was combined with linear discriminative analysis (LDA) techniques to produce a new face recognition methodology. The contribution of this paper is generated of one-dimensional face feature set by LDA from original image database to training of 1D-DCNN classifier, that it contributed in the improvement of facial recognition performance. The model has been tested on MCUT dataset consisted of 3755 images for 276 classes. The results of the implementation of face recognition were accuracy of 100%, precision of 100%, recall of 100%, and F-measure of 100%.

Attentional modulation of neural dynamics in tactile perception of complex regional pain syndrome patients.

Body perceptual disturbances are an increasingly acknowledged set of symptoms and possible clinical markers of complex regional pain syndrome (CRPS), but the neurophysiological and neurocognitive changes that underlie them are still far from being clear. We adopted a multivariate and neurodynamical approach to the analysis of EEG modulations evoked by touch to highlight differences between patients and healthy controls, between affected and unaffected side of the body, and between "passive" (i.e., no task demands and equiprobable digit stimulation) and "active" tactile processing (i.e., where a digit discrimination task was administered and spatial probability manipulated). When correct identifications are considered, an early reduction in cortical decodability (28-56 ms) distinguishes CRPS patients from healthy volunteers. However, when error trials are included in the classifier's training, there is an unexpected increased decodability in the CRPS group compared with healthy volunteers (280-320 ms). These group differences in neural processing seemed to be driven by the affected rather than the unaffected side. We corroborated these findings with several exploratory analyses of neural representation dynamics and behavioural modelling, highlighting the need for single participant analyses. Although several limitations impacted the robustness and generalizability of these comparisons, the proposed analytical approach yielded promising insights (as well as possible biomarkers based on neural dynamics) into the relatively unexplored alterations of tactile decision-making and attentional control mechanisms in chronic CRPS.

Robust semi-supervised classification based on data augmented online ELMs with deep features

One important strategy in semi-supervised learning is to utilize the predicted pseudo labels of unlabeled data to relieve the overdependence on the ground truth of supervised learning algorithms. However, the performance of such kinds of semi-supervised methods heavily relies on the quality of pseudo labels. To address this issue, a robust semi-supervised classification method, named data augmented online extreme learning machines (ELMs) with deep features (DF-DAELM) is proposed. This method firstly extracts features and infers labels for unlabeled data through self-training. Then, with the learned features and inferred labels, two noise-robust shallow classifiers based on data augmentation (i.e., SLI-OELM and CR-OELM) are proposed to eliminate the adverse effects of noises on classifier training. Specifically, inspired by label smoothing, a data augmented method, SLI-OELM is designed based on stochastic linear interpolation to improve the robustness of classifiers based on ELMs. Furthermore, based on the smoothing assumption, the proposed CR-OELM utilizes an ℓ2-norm consistency regularization term to implicitly weight noisy samples. Comprehensive experiments demonstrate that DF-DAELM achieves competitive or even better performance on CIFAR-10/100 and SVHN over the related state-of-the-art methods. Meanwhile, for the proposed classifiers, experimental results on the MNIST dataset with different noise levels and sample scales demonstrate their superior performance, especially when the sample scale is small (≤20K) and the noise is strong (40%∼80%).

Playing to distraction: towards a robust training of CNN classifiers through visual explanation techniques

The field of deep learning is evolving in different directions, with still the need for more efficient training strategies. In this work, we present a novel and robust training scheme that integrates visual explanation techniques in the learning process. Unlike the attention mechanisms that focus on the relevant parts of images, we aim to improve the robustness of the model by making it pay attention to other regions as well. Broadly speaking, the idea is to distract the classifier in the learning process to force it to focus not only on relevant regions but also on those that, a priori, are not so informative for the discrimination of the class. We tested the proposed approach by embedding it into the learning process of a convolutional neural network for the analysis and classification of two well-known datasets, namely Stanford cars and FGVC-Aircraft. Furthermore, we evaluated our model on a real-case scenario for the classification of egocentric images, allowing us to obtain relevant information about peoples' lifestyles. In particular, we work on the challenging EgoFoodPlaces dataset, achieving state-of-the-art results with a lower level of complexity. The obtained results indicate the suitability of our proposed training scheme for image classification, improving the robustness of the final model.

A patch-based and multi-instance learning strategy for pneumothorax classification on chest X-rays

Pneumothorax is a lung emergency. Automated computer-aid pneumothorax diagnosis based on chest X-ray can help reduce the diagnostic time and save valuable time for the treatment. A total of 21,759 patient’s frontal-view chest X-ray images from one medical center are used in this study. The dataset is divided into two categories: pneumothorax and non-pneumothorax, which are evaluated by two radiologists with over ten years of practical experience. A two-stage training for pneumothorax classification based on multi-instance learning (MIL) are proposed, first training a patch-level classifier, followed by an image-level classifier training, which is initialized with the patch pre-trained weights. The image-level classifier initialized with patch pre-trained weights achieves good classification performance with the F1-score, accuracy and recall of 0.869, 0.915 and 0.843 respectively, which are larger compared to that of the model initialized without patch pre-trained weights (0.785, 0.878 and 0.783). The two-stage training strategy can improve the performance of pneumothorax classification and does not require too high GPU memory and long training time.

The 30-days hospital readmission risk in diabetic patients: predictive modeling with machine learning classifiers

Background and objectivesDiabetes mellitus is a major chronic disease that results in readmissions due to poor disease control. Here we established and compared machine learning (ML)-based readmission prediction methods to predict readmission risks of diabetic patients.MethodsThe dataset analyzed in this study was acquired from the Health Facts Database, which includes over 100,000 records of diabetic patients from 1999 to 2008. The basic data distribution characteristics of this dataset were summarized and then analyzed. In this study, 30-days readmission was defined as a readmission period of less than 30 days. After data preprocessing and normalization, multiple risk factors in the dataset were examined for classifier training to predict the probability of readmission using ML models. Different ML classifiers such as random forest, Naive Bayes, and decision tree ensemble were adopted to improve the clinical efficiency of the classification. In this study, the Konstanz Information Miner platform was used to preprocess and model the data, and the performances of the different classifiers were compared.ResultsA total of 100,244 records were included in the model construction after the data preprocessing and normalization. A total of 23 attributes, including race, sex, age, admission type, admission location, length of stay, and drug use, were finally identified as modeling risk factors. Comparison of the performance indexes of the three algorithms revealed that the RF model had the best performance with a higher area under receiver operating characteristic curve (AUC) than the other two algorithms, suggesting that its use is more suitable for making readmission predictions.ConclusionThe factors influencing 30-days readmission predictions in diabetic patients, including number of inpatient admissions, age, diagnosis, number of emergencies, and sex, would help healthcare providers to identify patients who are at high risk of short-term readmission and reduce the probability of 30-days readmission. The RF algorithm with the highest AUC is more suitable for making 30-days readmission predictions and deserves further validation in clinical trials.

Comparing Three Machine Learning Techniques for Building Extraction from a Digital Surface Model

Automatic building extraction from high-resolution remotely sensed data is a major area of interest for an extensive range of fields (e.g., urban planning, environmental risk management) but challenging due to urban morphology complexity. Among the different methods proposed, the approaches based on supervised machine learning (ML) achieve the best results. This paper aims to investigate building footprint extraction using only high-resolution raster digital surface model (DSM) data by comparing the performance of three different popular supervised ML models on a benchmark dataset. The first two methods rely on a histogram of oriented gradients (HOG) feature descriptor and a classical ML (support vector machine (SVM)) or a shallow neural network (extreme learning machine (ELM)) classifier, and the third model is a fully convolutional network (FCN) based on deep learning with transfer learning. Used data were obtained from the International Society for Photogrammetry and Remote Sensing (ISPRS) and cover the urban areas of Vaihingen an der Enz, Potsdam, and Toronto. The results indicated that performances of models based on shallow ML (feature extraction and classifier training) are affected by the urban context investigated (F1 scores from 0.49 to 0.81), whereas the FCN-based model proved to be the most robust and best-performing method for building extraction from a high-resolution raster DSM (F1 scores from 0.80 to 0.86).

PERFORMANCE OF THE SUPPORT VECTOR MACHINE AND ARTIFICIAL NEURAL NETWORK CLASSIFIERS FOR ROADS IDENTIFICATION

Abstract. The objective of this project was to compare two non-parametric classification methods (“Support Vector Machine” – SVM and “Artificial Neural Networks” – ANN) of road regions in high spatial resolution images and associated with data from Airborne Laser Scanning. The study aims to verify what kind of influence the layers of attributes have on the performance from respective classifiers: SVM and RNA. Our method based on tests of this classifiers on 4 bands of airborne images and normalization of the digital surface model (DSM) for showing only information on objects height in relation to ground and not of these in relation to the ground and relief, generating band 5. The samples were used to train chosen non-parametric classifiers (training sets for each different input image/landscape). All classifications had the same set of training samples and the same classification parameters. The optimal parameters for classifications were obtained through the existing library in the Weka mining package: LibSVM and LibMultiLayerPerceptron. Our results demonstrated the existence of a direct relationship between the elevation band of the targets in relation to the terrain (band 05) with the improvement of their performance and lower degree of between bands correlation can also be considered a factor that has a positive influence. As for Neural Networks, the experiment results demonstrate that the presence of the near infrared band (band 04) was decisive for the performance improving of certain combinations in relation to others.

A High Accuracy Electrographic Seizure Classifier Trained Using Semi-Supervised Labeling Applied to a Large Spectrogram Dataset.

The objective of this study was to explore using ECoG spectrogram images for training reliable cross-patient electrographic seizure classifiers, and to characterize the classifiers’ test accuracy as a function of amount of training data. ECoG channels in ∼138,000 time-series ECoG records from 113 patients were converted to RGB spectrogram images. Using an unsupervised spectrogram image clustering technique, manual labeling of 138,000 ECoG records (each with up to 4 ECoG channels) was completed in 320 h, which is an estimated 5 times faster than manual labeling without ECoG clustering. For training supervised classifier models, five random folds of data were created; with each fold containing 72, 18, and 23 patients’ data for model training, validation and testing respectively. Five convolutional neural network (CNN) architectures, including two with residual connections, were trained. Cross-patient classification accuracies and F1 scores improved with model complexity, with the shallowest 6-layer model (with ∼1.5 million trainable parameters) producing a class-balanced seizure/non-seizure classification accuracy of 87.9% on ECoG channels and the deepest ResNet50-based model (with ∼23.5 million trainable parameters) producing a classification accuracy of 95.7%. The trained ResNet50-based model additionally had 93.5% agreement in scores with an independent expert labeller. Visual inspection of gradient-based saliency maps confirmed that the models’ classifications were based on relevant portions of the spectrogram images. Further, by repeating training experiments with data from varying number of patients, it was found that ECoG spectrogram images from just 10 patients were sufficient to train ResNet50-based models with 88% cross-patient accuracy, while at least 30 patients’ data was required to produce cross-patient classification accuracies of >90%.

AN UNSUPERVISED METHOD BASED ON FIRE INDEX ENHANCEMENT AND GRNN FOR AUTOMATED BURNED AREA MAPPING FROM SINGLE-PERIOD REMOTE SENSING IMAGERY

Abstract. In the current situation of frequent forest fires, the study of forest burned area mapping is important. However, there is still room for improvement in the accuracy of existing forest burning area mapping methods. Therefore, in this paper, an unsupervised method based on fire index enhancement and GRNN (General Regression Neural Network) is proposed for automated forest burned area mapping from single-date post-fire remote sensing imagery. The proposed method first uses adaptive spatial context information to enhance the generated fire index to improve its ability to indicate the burned areas. Then the uncertainty analysis is performed on the enhanced fire index to extract reliable burned samples and non-burned samples for subsequent classifier training. Finally, the improved GRNN model considering the spatial correlation of pixels is used as a classifier to binarize the enhanced fire index to generate the final burned area map. Based on two commonly used fire indexes, NBR (Normalized Burn Ratio) and BAI (Burned Area Index), this paper conducts burned area mapping experiments on a post-fire image of a forest area in Inner Mongolia, China to test the effectiveness of the proposed method, and two commonly used threshold methods (Otsu and Kmeans clustering) are also used to conduct burned area mapping based on threshold segmentation of fire index for comparison experiments. The experimental results prove the effectiveness and superiority of the proposed method. The proposed method is unsupervised and automated, so it has high application value and potential under the current situation of frequent forest fires.

MI-MOTE: Multiple imputation-based minority oversampling technique for imbalanced and incomplete data classification

Class imbalance and data incompleteness problems occur simultaneously in many real-world classification datasets, which negatively affects the training of classifiers. Given an imbalanced and incomplete training dataset, the conventional approach is to address these two problems sequentially by handling data incompleteness first and then focusing on class imbalance. In this study, we propose a multiple imputation-based minority oversampling technique, named MI-MOTE, to address imbalanced and incomplete data classification simultaneously. Majority instances are imputed once and minority instances are oversampled using multiple different imputations without directly manipulating any of their observed values. Accordingly, minority instances are diversified with less data distortion compared to the conventional approach. The proposed method is applied in the data preprocessing phase, meaning it can be used with any type of classifier. Experimental results for benchmark datasets with various missing rates demonstrate the effectiveness of the proposed method.

Application of noninvasive magnetomyography in labour imminency prediction for term and preterm pregnancies and ethnicity specific labour prediction

This paper investigates the application of magnetomyography (MMG) signals from uterine contractions in pregnant patients towards the prediction of labour imminency within or above 48 h. The study utilised the MMG signals collected from a host of pregnant patients retrieved from a Physionet database, which also contained information regarding patients’ ethnicity and pregnancy. Utilising​ the information available in addition to the dataset, the study investigated the prospect of designing an ethnic specific labour imminency classifier to allow for an enhanced prediction, with an emphasis on Black and Caucasian ethnicities due to the nature of the data. Using an extended feature vector and a support vector machine (SVM) classifier, it was seen that the labour imminency was enhanced across the various classifier metrics considered in the ethnic specific classifier when compared with the generalised classifier. The results from the classification exercise, which considered the fusion of MMG signal information with the information on patients’ records, showed greater variability and a slightly lower classifier performance, thus suggesting that the MMG signals present a more reliable way of classifier training. Subsequent work in this area would now involve the application of optimisation algorithms to select an optimal number of electrodes that can be used for data acquisition, and thereby contributing towards the lowering of the cost associated with the implementation of the method using the MMG instrumentation.

Online control of an assistive active glove by slow cortical signals in patients with amyotrophic lateral sclerosis

Objective. A brain–computer interface (BCI) allows users to control external devices using brain signals that can be recorded non-invasively via electroencephalography (EEG). Movement related cortical potentials (MRCPs) are an attractive option for BCI control since they arise naturally during movement execution and imagination, and therefore, do not require an extensive training. This study tested the feasibility of online detection of reaching and grasping using MRCPs for the application in patients suffering from amyotrophic lateral sclerosis (ALS). Approach. A BCI system was developed to trigger closing of a soft assistive glove by detecting a reaching movement. The custom-made software application included data collection, a novel method for collecting the input data for classifier training from the offline recordings based on a sliding window approach, and online control of the glove. Eight healthy subjects and two ALS patients were recruited to test the developed BCI system. They performed assessment blocks without the glove active (NG), in which the movement detection was indicated by a sound feedback, and blocks (G) in which the glove was controlled by the BCI system. The true positive rate (TPR) and the positive predictive value (PPV) were adopted as the outcome measures. Correlation analysis between forehead EEG detecting ocular artifacts and sensorimotor area EEG was conducted to confirm the validity of the results. Main results. The overall median TPR and PPV were >0.75 for online BCI control, in both healthy individuals and patients, with no significant difference across the blocks (NG versus G). Significance. The results demonstrate that cortical activity during reaching can be detected and used to control an external system with a limited amount of training data (30 trials). The developed BCI system can be used to provide grasping assistance to ALS patients.

АЛГОРИТМ ОБНАРУЖЕНИЯ ОБЪЕКТОВ ДЛЯ ОПТИКО-ЭЛЕКТРОННЫХ СИСТЕМ С ОБУЧЕНИЕМ В РЕАЛЬНОМ ВРЕМЕНИ

In order to create a new algorithm for automatic detection of objects with real-time training, astudy of the world scientific groundwork in the field of general-purpose automatic tracking with theability to recognize a tracked object with the potential for application in embedded computing systemsof optoelectronic systems of promising robotic complexes was carried out. Based on the conductedresearch, methods and approaches were selected and tested that allow, with the greatest accuracy,while maintaining high computational efficiency, to provide training of classifiers on the fly(online learning) without a priori knowledge of the type of tracking object and to ensure the subsequentdetection of the original object in the event of its short-term loss. Such methods include a histogramof oriented gradients – a descriptor of key features based on the analysis of the distribution ofthe brightness gradients of the object image. Its use allows you to reduce the amount of informationused without losing key data about the object and to increase the speed of image processing. Thearticle substantiates the choice of one of the real-time classification algorithms that allows solvingthe problem of binary classification – the support vector machine. Due to the high speed of data processingand the need for a small amount of initial training data to construct a separating hyperplane,on the basis of which the classification of objects is done, this method is chosen as the most suitablefor solving the problem. For online training, a modification of the support vector machine methodwas chosen, which implements stochastic gradient descent at each step of the algorithm – Pegasos.The authors of the study carried out the development and semi-natural modeling of the selected algorithm,evaluated the effectiveness of its work in the tasks of detecting an object of interest in real timewith preliminary online training in the process of tracking the object. The developed algorithm hasshown high efficiency in solving the problem and is planned to be implemented as part of a specialsoftware for optoelectronic systems of advanced robotic systems. In the conclusion, proposals arepresented to further improve the accuracy and probability of the object detection by the developedalgorithm, as well as for improving its performance by optimizing calculations.

Robust North Atlantic right whale detection using deep learning models for denoising.

This paper proposes a robust system for detecting North Atlantic right whales by using deep learning methods to denoise noisy recordings. Passive acoustic recordings of right whale vocalisations are subject to noise contamination from many sources, such as shipping and offshore activities. When such data are applied to uncompensated classifiers, accuracy falls substantially. To build robustness into the detection process, two separate approaches that have proved successful for image denoising are considered. Specifically, a denoising convolutional neural network and a denoising autoencoder, each of which is applied to spectrogram representations of the noisy audio signal, are developed. Performance is improved further by matching the classifier training to include the vestigial signal that remains in clean estimates after the denoising process. Evaluations are performed first by adding white, tanker, trawler, and shot noises at signal-to-noise ratios from -10 to +5 dB to clean recordings to simulate noisy conditions. Experiments show that denoising gives substantial improvements to accuracy, particularly when using the vestigial-trained classifier. A final test applies the proposed methods to previously unseen noisy right whale recordings and finds that denoising is able to improve performance over the baseline clean-trained model in this new noise environment.

Building an Open Source Classifier for the Neonatal EEG Background: A Systematic Feature-Based Approach From Expert Scoring to Clinical Visualization.

Neonatal brain monitoring in the neonatal intensive care units (NICU) requires a continuous review of the spontaneous cortical activity, i.e., the electroencephalograph (EEG) background activity. This needs development of bedside methods for an automated assessment of the EEG background activity. In this paper, we present development of the key components of a neonatal EEG background classifier, starting from the visual background scoring to classifier design, and finally to possible bedside visualization of the classifier results. A dataset with 13,200 5-minute EEG epochs (8–16 channels) from 27 infants with birth asphyxia was used for classifier training after scoring by two independent experts. We tested three classifier designs based on 98 computational features, and their performance was assessed with respect to scoring system, pre- and post-processing of labels and outputs, choice of channels, and visualization in monitor displays. The optimal solution achieved an overall classification accuracy of 97% with a range across subjects of 81–100%. We identified a set of 23 features that make the classifier highly robust to the choice of channels and missing data due to artefact rejection. Our results showed that an automated bedside classifier of EEG background is achievable, and we publish the full classifier algorithm to allow further clinical replication and validation studies.

Current and future deep learning algorithms for tandem mass spectrometry (MS/MS)-based small molecule structure elucidation.

Structure elucidation of small molecules has been one of the cornerstone applications of mass spectrometry for decades. Despite the increasing availability of software tools, structure elucidation from tandem mass spectrometry (MS/MS) data remains a challenging task, leaving many spectra unidentified. However, as an increasing number of reference MS/MS spectra are being curated at a repository scale and shared on public servers, there is an exciting opportunity to develop powerful new deep learning (DL) models for automated structure elucidation. Recent early-stage DL frameworks mostly follow a "two-step approach" that translates MS/MS spectra to database structures after first predicting molecular descriptors. The related architectures could suffer from: (1) computational complexity because of the separate training of descriptor-specific classifiers, (2) the high dimensional nature of mass spectral data and information loss due to data preprocessing, (3) low substructure coverage and class imbalance problem of predefined molecular fingerprints. Inspired by successful DL frameworks employed in drug discovery fields, we have conceptualized and designed hypothetical DL architectures to tackle the above issues. For (1), we recommend multitask learning to achieve better performance with fewer classifiers by grouping structurally related descriptors. For (2) and (3), we introduce feature engineering to extract condensed and higher-order information from spectra and structure data. For instance, encoding spectra with subtrees and pre-calculated spectral patterns add peak interactions to the model input. Encoding structures with graph convolutional networks incorporates connectivity within a molecule. The joint embedding of spectra and structures can enable simultaneous spectral library and molecular database search. In principle, given enough training data, adapted DL architectures, optimal hyperparameters and computing power, DL frameworks can predict small molecule structures, completely or at least partially, from MS/MS spectra. However, their performance and general applicability should be fairly evaluated against classical machine learning frameworks.

A novel antibacterial peptide recognition algorithm based on BERT.

As the best substitute for antibiotics, antimicrobial peptides (AMPs) have important research significance. Due to the high cost and difficulty of experimental methods for identifying AMPs, more and more researches are focused on using computational methods to solve this problem. Most of the existing calculation methods can identify AMPs through the sequence itself, but there is still room for improvement in recognition accuracy, and there is a problem that the constructed model cannot be universal in each dataset. The pre-training strategy has been applied to many tasks in natural language processing (NLP) and has achieved gratifying results. It also has great application prospects in the field of AMP recognition and prediction. In this paper, we apply the pre-training strategy to the model training of AMP classifiers and propose a novel recognition algorithm. Our model is constructed based on the BERT model, pre-trained with the protein data from UniProt, and then fine-tuned and evaluated on six AMP datasets with large differences. Our model is superior to the existing methods and achieves the goal of accurate identification of datasets with small sample size. We try different word segmentation methods for peptide chains and prove the influence of pre-training steps and balancing datasets on the recognition effect. We find that pre-training on a large number of diverse AMP data, followed by fine-tuning on new data, is beneficial for capturing both new data's specific features and common features between AMP sequences. Finally, we construct a new AMP dataset, on which we train a general AMP recognition model.

Efficient and accurate identification of ear diseases using an ensemble deep learning model

Early detection and appropriate medical treatment are of great use for ear disease. However, a new diagnostic strategy is necessary for the absence of experts and relatively low diagnostic accuracy, in which deep learning plays an important role. This paper puts forward a mechanic learning model which uses abundant otoscope image data gained in clinical cases to achieve an automatic diagnosis of ear diseases in real time. A total of 20,542 endoscopic images were employed to train nine common deep convolution neural networks. According to the characteristics of the eardrum and external auditory canal, eight kinds of ear diseases were classified, involving the majority of ear diseases, such as normal, Cholestestoma of the middle ear, Chronic suppurative otitis media, External auditory cana bleeding, Impacted cerumen, Otomycosis external, Secretory otitis media, Tympanic membrane calcification. After we evaluate these optimization schemes, two best performance models are selected to combine the ensemble classifiers with real-time automatic classification. Based on accuracy and training time, we choose a transferring learning model based on DensNet-BC169 and DensNet-BC1615, getting a result that each model has obvious improvement by using these two ensemble classifiers, and has an average accuracy of 95.59%. Considering the dependence of classifier performance on data size in transfer learning, we evaluate the high accuracy of the current model that can be attributed to large databases. Current studies are unparalleled regarding disease diversity and diagnostic precision. The real-time classifier trains the data under different acquisition conditions, which is suitable for real cases. According to this study, in the clinical case, the deep learning model is of great use in the early detection and remedy of ear diseases.

Comparison of machine-learning methodologies for accurate diagnosis of sepsis using microarray gene expression data.

We investigate the feasibility of molecular-level sample classification of sepsis using microarray gene expression data merged by in silico meta-analysis. Publicly available data series were extracted from NCBI Gene Expression Omnibus and EMBL-EBI ArrayExpress to create a comprehensive meta-analysis microarray expression set (meta-expression set). Measurements had to be obtained via microarray-technique from whole blood samples of adult or pediatric patients with sepsis diagnosed based on international consensus definition immediately after admission to the intensive care unit. We aggregate trauma patients, systemic inflammatory response syndrome (SIRS) patients, and healthy controls in a non-septic entity. Differential expression (DE) analysis is compared with machine-learning-based solutions like decision tree (DT), random forest (RF), support vector machine (SVM), and deep-learning neural networks (DNNs). We evaluated classifier training and discrimination performance in 100 independent iterations. To test diagnostic resilience, we gradually degraded expression data in multiple levels. Clustering of expression values based on DE genes results in partial identification of sepsis samples. In contrast, RF, SVM, and DNN provide excellent diagnostic performance measured in terms of accuracy and area under the curve (>0.96 and >0.99, respectively). We prove DNNs as the most resilient methodology, virtually unaffected by targeted removal of DE genes. By surpassing most other published solutions, the presented approach substantially augments current diagnostic capability in intensive care medicine.