Combination Of Linear Discriminant Analysis Research Articles

Somnotate: A probabilistic sleep stage classifier for studying vigilance state transitions.

Electrophysiological recordings from freely behaving animals are a widespread and powerful mode of investigation in sleep research. These recordings generate large amounts of data that require sleep stage annotation (polysomnography), in which the data is parcellated according to three vigilance states: awake, rapid eye movement (REM) sleep, and non-REM (NREM) sleep. Manual and current computational annotation methods ignore intermediate states because the classification features become ambiguous, even though intermediate states contain important information regarding vigilance state dynamics. To address this problem, we have developed "Somnotate"-a probabilistic classifier based on a combination of linear discriminant analysis (LDA) with a hidden Markov model (HMM). First we demonstrate that Somnotate sets new standards in polysomnography, exhibiting annotation accuracies that exceed human experts on mouse electrophysiological data, remarkable robustness to errors in the training data, compatibility with different recording configurations, and an ability to maintain high accuracy during experimental interventions. However, the key feature of Somnotate is that it quantifies and reports the certainty of its annotations. We leverage this feature to reveal that many intermediate vigilance states cluster around state transitions, whereas others correspond to failed attempts to transition. This enables us to show for the first time that the success rates of different types of transition are differentially affected by experimental manipulations and can explain previously observed sleep patterns. Somnotate is open-source and has the potential to both facilitate the study of sleep stage transitions and offer new insights into the mechanisms underlying sleep-wake dynamics.

Raman spectroscopy and Raman optical activity of blood plasma for differential diagnosis of gastrointestinal cancers

Improving the early diagnosis of gastrointestinal cancers is a crucial step in reducing their mortality. Given the non-specificity of the initial symptoms, the ability of any diagnostic method to differentiate between various types of gastrointestinal cancers also needs to be addressed. To detect disease-specific alterations in biomolecular structure and composition of the blood plasma, we have implemented an approach combining Raman spectroscopy and its conformation-sensitive polarized version, Raman optical activity, to analyze blood plasma samples of patients suffering from three different types of gastrointestinal cancer – hepatocellular, colorectal and pancreatic. First, we aimed to discriminate any type of gastrointestinal cancer from healthy control individuals; inthenext step, the focus was on differentiating among the three cancer types studied. The more straightforward of the two statistical approaches tested, the combination of linear discriminant analysis and principal component analysis applied to the entire spectral dataset, allowed the discrimination of cancer and control samples with 87% accuracy. The three gastrointestinal cancers were classified with an overall accuracy of 76%. The second method, the linear discriminant analysis applied to a selection of spectral bands, yielded even higher values. Cancer and control samples were distinguished with 89% accuracy and hepatocellular, colorectal and pancreatic cancer with an overall accuracy of 87%. The results obtained in our study suggest that the proposed approach may become a disease-specific diagnostic tool in daily clinical practice.

Impact of a resistance gene against a fungal pathogen on the plant host residue microbiome: The case of the Leptosphaeria maculans–Brassica napus pathosystem

Oilseed rape residues are a crucial determinant of stem canker epidemiology as they support the sexual reproduction of the fungal pathogen Leptosphaeria maculans. The aim of this study was to characterize the impact of a resistance gene against L. maculans infection on residue microbial communities and to identify microorganisms interacting with this pathogen during residue degradation. We used near‐isogenic lines to obtain healthy and infected host plants. The microbiome associated with the two types of plant residues was characterized by metabarcoding. A combination of linear discriminant analysis and ecological network analysis was used to compare the microbial communities and to identify microorganisms interacting with L. maculans. Fungal community structure differed between the two lines at harvest, but not subsequently, suggesting that the presence/absence of the resistance gene influences the microbiome at the base of the stem whilst the plant is alive, but that this does not necessarily lead to differential colonization of the residues by fungi. Direct interactions with other members of the community involved many fungal and bacterial amplicon sequence variants (ASVs). L. maculans appeared to play a minor role in networks, whereas one ASV affiliated to Plenodomus biglobosus (synonym Leptosphaeria biglobosa) from the Leptosphaeria species complex may be considered a keystone taxon in the networks at harvest. This approach could be used to identify and promote microorganisms with beneficial effects against residue‐borne pathogens and, more broadly, to decipher the complex interactions between multispecies pathosystems and other microbial components in crop residues.

Human body activity recognition using wearable inertial sensors integrated with a feature extraction–based machine-learning classification algorithm

This study employed wearable inertial sensors integrated with an activity-recognition algorithm to recognize six types of daily activities performed by humans, namely walking, ascending stairs, descending stairs, sitting, standing, and lying. The sensor system consisted of a microcontroller, a three-axis accelerometer, and a three-axis gyro; the algorithm involved collecting and normalizing the activity signals. To simplify the calculation process and to maximize the recognition accuracy, the data were preprocessed through linear discriminant analysis; this reduced their dimensionality and captured their features, thereby reducing the feature space of the accelerometer and gyro signals; they were then verified through the use of six classification algorithms. The new contribution is that after feature extraction, data classification results indicated that an artificial neural network was the most stable and effective of the six algorithms. In the experiment, 20 participants equipped the wearable sensors on their waists to record the aforementioned six types of daily activities and to verify the effectiveness of the sensors. According to the cross-validation results, the combination of linear discriminant analysis and an artificial neural network was the most stable classification algorithm for data generalization; its activity-recognition accuracy was 87.37% on the training data and 80.96% on the test data.

Combination of linear discriminant analysis and expert opinion for the construction of credit rating models: The case of SMEs

The construction of an internal rating model is the main task for the bank in the framework of the IRB-foundation approach the fact that it is necessary to determine the probability of default by r...

Optimised deep learning features for improved melanoma detection

In this article, we are addressing the question of effective usage of the feature set extracted from deep learning models pre-trained on ImageNet. Exploring this option will offer very fast and attractive alternative to transfer learning strategies. The traditional task of skin lesion recognition consists of several stages, where the automated system is typically trained on preprocessed images with known diagnosis, which allows classification of new samples to predefined categories. For this task, we are proposing here an improved melanoma detection method based on the combination of linear discriminant analysis (LDA) and the features extracted from the deep learning approach. We are examining the usage of the LDA approach on activation of the fully-connected layer of deep learning in order to increase the classification accuracy and at the same time to reduce the feature space dimensionality. We tested our method on five different classifiers and evaluated results using various metrics. The presented comparison demonstrates the very high effectiveness of the suggested feature reduction, which leads not only to the significant lowering of employed features but also to the increasing performance of all tested classifiers in almost all measured characteristics.

부정감성 인식을 위한 생체신호 기반의 특징 선택 알고리즘 개발

감성은 인간의 삶과 밀접한 관련을 가지고 있으며 이는 집중력, 학습능력 등 많은 부분에 영향을 주어 다양한 행동 패턴을 가지게 한다. 따라서 본 논문의 목적은 부정감성을 구분하기 위하여 생체신호를 기반으로 주요한 특징들을 추출하는 것이다. 이를 위해 본 논문에서는 심전도, 뇌파, 피부 온도와 피부전도도를 기반으로 생체신호를 측정한 후, 선형분류기와 유전 알고리즘의 조합으로 정확하고 신속한 알고리즘 개발하고, 주요 특징을 추출하였다. 그 결과, 알고리즘은 최대 96.4%의 정확도를 가짐을 확인할 수 있었고, 추출된 파라미터는 심박변이도의 Mean, RMSSD, NN50과 뇌파의 전두엽 영역에서의 <TEX>${\sigma}$</TEX>파와 <TEX>${\alpha}$</TEX>파의 주파수 파워, 두정엽 영역에서 <TEX>${\alpha}$</TEX>파, <TEX>${\beta}$</TEX>파, <TEX>${\gamma}$</TEX>파와 의 주파수 파워, 그리고 피부온도의 평균과 표준편차 값이었다. 이에 따라 각 각의 생체신호를 기반으로 한 추출 된 특징들은 부정감성의 분류에 있어 중요한 역할을 함을 확인할 수 있었다. Emotion is closely related to the life of human, so has effect on many parts such as concentration, learning ability, etc. and makes to have different behavior patterns. The purpose of this paper is to extract important features based on physiological signals to recognize negative emotion. In this paper, after acquisition of electrocardiography(ECG), electroencephalography(EEG), skin temperature(SKT) and galvanic skin response(GSR) measurements based on physiological signals, we designed an accurate and fast algorithm using combination of linear discriminant analysis(LDA) and genetic algorithm(GA), then we selected important features. As a result, the accuracy of the algorithm is up to 96.4% and selected features are Mean, root mean square successive difference(RMSSD), NN intervals differing more than 50ms(NN50) of heart rate variability(HRV), <TEX>${\sigma}$</TEX> and <TEX>${\alpha}$</TEX> frequency power of EEG from frontal region, <TEX>${\alpha}$</TEX>, <TEX>${\beta}$</TEX>, and <TEX>${\gamma}$</TEX> frequency power of EEG from central region, and mean and standard deviation of SKT. Therefore, the features play an important role to recognize negative emotion.

Exploring the knowledge contained in neuroimages: Statistical discriminant analysis and automatic segmentation of the most significant changes

The aim of this article is to propose an integrated framework for extracting and describing patterns of disorders from medical images using a combination of linear discriminant analysis and active contour models. A multivariate statistical methodology was first used to identify the most discriminating hyperplane separating two groups of images (from healthy controls and patients with schizophrenia) contained in the input data. After this, the present work makes explicit the differences found by the multivariate statistical method by subtracting the discriminant models of controls and patients, weighted by the pooled variance between the two groups. A variational level-set technique was used to segment clusters of these differences. We obtain a label of each anatomical change using the Talairach atlas. In this work all the data was analysed simultaneously rather than assuming a priori regions of interest. As a consequence of this, by using active contour models, we were able to obtain regions of interest that were emergent from the data. The results were evaluated using, as gold standard, well-known facts about the neuroanatomical changes related to schizophrenia. Most of the items in the gold standard was covered in our result set. We argue that such investigation provides a suitable framework for characterising the high complexity of magnetic resonance images in schizophrenia as the results obtained indicate a high sensitivity rate with respect to the gold standard.

Multiclass Classification Based on Extended Support Vector Data Description

We propose two variations of the support vector data description (SVDD) with negative samples (NSVDD) that learn a closed spherically shaped boundary around a set of samples in the target class by involving different forms of slack vectors, including the two-norm NSVDD and nu-NSVDD. We extend the NSVDDs to solve the multiclass classification problems based on the distances between the samples and the centers of the learned spherically shaped boundaries in a kernel-defined feature space by using a combination of linear discriminant analysis (LDA) and nearest-neighbor (NN) rule. Extensive simulations are developed with one real-world data set on the automatic monitoring of roller bearings with vibration signals and eight benchmark data sets for both binary and multiclass classification. The benchmark testing results show that our proposed methods provide lower classification error rates and smaller standard deviations with the cross-validation procedure. The two-norm NSVDD with the LDA-NN rule recorded a test accuracy of 100.0% for the binary fault detection of roller bearings and 99.9% for the multiclass classification of roller bearings under six conditions.

Study of different kinds of “Pesto Genovese” by the analysis of their volatile fraction and chemometric methods

The volatile fraction of different kinds of Pesto genovese, a typical Italian basil-based pasta sauce, was analysed by Headspace Sorptive Extraction (HSSE)-Gas Chromatography–Mass Spectrometry (GC–MS) coupled with chemometric methods. Three kinds of commercial samples were considered: (a) non heat-processed shop manufactured, (b) heat-processed industrial and (c) non heat-processed industrial samples. The category of heat-processed samples was easily discriminated by the amounts of the terpene hydrocarbon α-terpinene. In order to distinguish the three categories, multivariate statistical analysis was then performed. Good prediction results were obtained by the combination of Linear Discriminant Analysis (LDA) and Stepwise LDA (STEPLDA): the percentage of correct predictions was 92 for food industry manufactured non heat-processed samples and 100 both for shop manufactured, non heat-processed samples, and for food industry manufactured, heat-processed samples.