The paper presents a study of two novel visual motion onset stimulus-based brain-computer interfaces (vmoBCI). Two settings are compared with afferent and efferent to a computer screen center motion patterns. Online vmoBCI experiments are conducted in an oddball event-related potential (ERP) paradigm allowing for "aha-responses" decoding in EEG brainwaves. A subsequent stepwise linear discriminant analysis classification (swLDA) classification accuracy comparison is discussed based on two inter-stimulus-interval (ISI) settings of 700 and 150 ms in two online vmoBCI applications with six and eight command settings. A research hypothesis of classification accuracy non-significant differences with various ISIs is confirmed based on the two settings of 700 ms and 150 ms, as well as with various numbers of ERP response averaging scenarios. The efferent in respect to display center visual motion patterns allowed for a faster interfacing and thus they are recommended as more suitable for the no-eye-movements requiring visual BCIs.
Structural relationships among proteins are important in the study of their evolution as well as in drug design and development. The protein 3D structure has been shown to be effective with respect to classifying proteins. Prior work has shown that the Double Centroid Reduced Representation (DCRR) model is a useful geometric representation for protein structure with respect to visual models, reducing the quantity of modeled information for each amino acid, yet retaining the most important geometrical and chemical features of each: the centroids of the backbone and of the side-chain. DCRR has not yet been applied in the calculation of geometric structural similarity. Meanwhile, multi-dimensional indexing (MDI) of protein structure combines protein structural analysis with distance metrics to facilitate structural similarity queries and is also used for clustering protein structures into related groups. In this respect, the combination of geometric models with MDI has been shown to be effective. Prior work, notably Distance and Density based Protein Indexing (DDPIn), applies MDI to protein models based on the geometry of the CA backbone. DDPIn distance metrics are based on radial and density functions that incorporate spherical-based metrics, and the indices are built from metric tree (M-tree) structures. This work combines DCRR with DDPIn for the development of new DCRR centroid-based metrics: spherical binning distance and inter-centroid spherical distance. The use of DCRR models will provide additional significant structural information via the inclusion of side-chain centroids. Additionally, the newly developed distance metric functions combined with DCRR and M-tree indexing attempt to improve upon the performance of prior work (DDPIn), given the same data set, with respect to both individual k-nearest neighbor (kNN) search queries as well as clustering all proteins in the index.
Despite the common misconception of nearly static organisms, plants do interact continuously with the environment and with each other. It is fair to assume that during their evolution they developed particular features to overcome problems and to exploit possibilities from environment. In this paper we introduce various quantitative measures based on recent advancements in complex network theory that allow to measure the effective similarities of various species. By using this approach on the similarity in fruit-typology ecological traits we obtain a clear plant classification in a way similar to traditional taxonomic classification. This result is not trivial, since a similar analysis done on the basis of diaspore morphological properties do not provide any clear parameter to classify plants species. Complex network theory can then be used in order to determine which feature amongst many can be used to distinguish scope and possibly evolution of plants. Future uses of this approach range from functional classification to quantitative determination of plant communities in nature.
Urban environments develop complex, non-obvious structures that are often hard to represent in the form of maps or guides. Finding the right place to go often requires intimate familiarity with the location in question and cannot easily be deduced by visitors. In this work, we exploit large-scale samples of usage information, in the form of mobile phone traces and geo-tagged Twitter messages in order to automatically explore and annotate city maps via kernel density estimation. Our experiments are based on one year's worth of mobile phone activity collected by Nokia's Mobile Data Challenge (MDC). We show that usage information can be a strong predictor of semantic place categories, allowing us to automatically annotate maps based on the behavior of the local user base.
Research is a continuous phenomenon. It is recursive in nature. Every research is based on some earlier research outcome. A general approach in reviewing the literature for a problem is to categorize earlier work for the same problem as positive and negative citations. In this paper, we propose a novel automated technique, which classifies whether an earlier work is cited as sentiment positive or sentiment negative. Our approach first extracted the portion of the cited text from citing paper. Using a sentiment lexicon we classify the citation as positive or negative by picking a window of at most five (5) sentences around the cited place (corpus). We have used Na\"ive-Bayes Classifier for sentiment analysis. The algorithm is evaluated on a manually annotated and class labelled collection of 150 research papers from the domain of computer science. Our preliminary results show an accuracy of 80%. We assert that our approach can be generalized to classification of scientific research papers in different disciplines.
The paper presents a study of an inter-stimulus interval (ISI) influence on a tactile point-pressure stimulus-based brain-computer interface's (tpBCI) classification accuracy. A novel tactile pressure generating tpBCI stimulator is also discussed, which is based on a three-by-three pins' matrix prototype. The six pin-linear patterns are presented to the user's palm during the online tpBCI experiments in an oddball style paradigm allowing for "the aha-responses" elucidation, within the event related potential (ERP). A subsequent classification accuracies' comparison is discussed based on two ISI settings in an online tpBCI application. A research hypothesis of classification accuracies' non-significant differences with various ISIs is confirmed based on the two settings of 120 ms and 300 ms, as well as with various numbers of ERP response averaging scenarios.
How does the human brain represent simple compositions of constituents: actors, verbs, objects, directions, and locations? Subjects viewed videos during neuroimaging (fMRI) sessions from which sentential descriptions of those videos were identified by decoding the brain representations based only on their fMRI activation patterns. Constituents (e.g., "fold" and "shirt") were independently decoded from a single presentation. Independent constituent classification was then compared to joint classification of aggregate concepts (e.g., "fold-shirt"); results were similar as measured by accuracy and correlation. The brain regions used for independent constituent classification are largely disjoint and largely cover those used for joint classification. This allows recovery of sentential descriptions of stimulus videos by composing the results of the independent constituent classifiers. Furthermore, classifiers trained on the words one set of subjects think of when watching a video can recognise sentences a different subject thinks of when watching a different video.
Most of the banks' operational risk internal models are based on loss pooling in risk and business line categories. The parameters and outputs of operational risk models are sensitive to the pooling of the data and the choice of the risk classification. In a simple model, we establish the link between the number of risk cells and the model parameters by requiring invariance of the bank's loss distribution upon a change in classification. We provide details on the impact of this requirement on the domain of attraction of the loss distribution, on diversification effects and on cell risk correlations.
Microblog classification has received a lot of attention in recent years. Different classification tasks have been investigated, most of them focusing on classifying microblogs into a small number of classes (five or less) using a training set of manually annotated tweets. Unfortunately, labelling data is tedious and expensive, and finding tweets that cover all the classes of interest is not always straightforward, especially when some of the classes do not frequently arise in practice. In this paper we study an approach to tweet classification based on distant supervision, whereby we automatically transfer labels from one social medium to another for a single-label multi-class classification task. In particular, we apply YouTube video classes to tweets linking to these videos. This provides for free a virtually unlimited number of labelled instances that can be used as training data. The classification experiments we have run show that training a tweet classifier via these automatically labelled data achieves substantially better performance than training the same classifier with a limited amount of manually labelled data; this is advantageous, given that the automatically labelled data come at no cost. Further investigation of our approach shows its robustness when applied with different numbers of classes and across different languages.
The typical process for classifying and submitting a newly sequenced virus to the NCBI database involves two steps. First, a BLAST search is performed to determine likely family candidates. That is followed by checking the candidate families with the Pairwise Sequence Alignment tool for similar species. The submitter's judgement is then used to determine the most likely species classification. The aim of this paper is to show that this process can be automated into a fast, accurate, one-step process using the proposed alignment-free method and properly implemented machine learning techniques. We present a new family of alignment-free vectorizations of the genome, the generalized vector, that maintains the speed of existing alignment-free methods while outperforming all available methods. This new alignment-free vectorization uses the frequency of genomic words (k-mers), as is done in the composition vector, and incorporates descriptive statistics of those k-mers' positional information, as inspired by the natural vector. We analyze 5 different characterizations of genome similarity using $k$-nearest neighbor classification, and evaluate these on two collections of viruses totaling over 10,000 viruses. We show that our proposed method performs better than, or as well as, other methods at every level of the phylogenetic hierarchy. The data and R code is available upon request.
Automated red blood cell (RBC) classification on blood smear images helps hematologists to analyze RBC lab results in a reduced time and cost. However, overlapping cells can cause incorrect predicted results, and so they have to be separated into multiple single RBCs before classifying. To classify multiple classes with deep learning, imbalance problems are common in medical imaging because normal samples are always higher than rare disease samples. This paper presents a new method to segment and classify RBCs from blood smear images, specifically to tackle cell overlapping and data imbalance problems. Focusing on overlapping cell separation, our segmentation process first estimates ellipses to represent RBCs. The method detects the concave points and then finds the ellipses using directed ellipse fitting. The accuracy from 20 blood smear images was 0.889. Classification requires balanced training datasets. However, some RBC types are rare. The imbalance ratio of this dataset was 34.538 for 12 RBC classes from 20,875 individual RBC samples. The use of machine learning for RBC classification with an imbalanced dataset is hence more challenging than many other applications. We analyzed techniques to deal with this problem. The best accuracy and F1-score were 0.921 and 0.8679, respectively, using EfficientNet-B1 with augmentation. Experimental results showed that the weight balancing technique with augmentation had the potential to deal with imbalance problems by improving the F1-score on minority classes, while data augmentation significantly improved the overall classification performance.
In this work we introduce attention as a state of the art mechanism for classification of radio galaxies using convolutional neural networks. We present an attention-based model that performs on par with previous classifiers while using more than 50% fewer parameters than the next smallest classic CNN application in this field. We demonstrate quantitatively how the selection of normalisation and aggregation methods used in attention-gating can affect the output of individual models, and show that the resulting attention maps can be used to interpret the classification choices made by the model. We observe that the salient regions identified by the our model align well with the regions an expert human classifier would attend to make equivalent classifications. We show that while the selection of normalisation and aggregation may only minimally affect the performance of individual models, it can significantly affect the interpretability of the respective attention maps and by selecting a model which aligns well with how astronomers classify radio sources by eye, a user can employ the model in a more effective manner.
Objective: To evaluate the impact on Electroencephalography (EEG) classification of different kinds of attention mechanisms in Deep Learning (DL) models. Methods: We compared three attention-enhanced DL models, the brand-new InstaGATs, an LSTM with attention and a CNN with attention. We used these models to classify normal and abnormal (i.e., artifactual or pathological) EEG patterns. Results: We achieved the state of the art in all classification problems, regardless the large variability of the datasets and the simple architecture of the attention-enhanced models. We could also prove that, depending on how the attention mechanism is applied and where the attention layer is located in the model, we can alternatively leverage the information contained in the time, frequency or space domain of the dataset. Conclusions: with this work, we shed light over the role of different attention mechanisms in the classification of normal and abnormal EEG patterns. Moreover, we discussed how they can exploit the intrinsic relationships in the temporal, frequency and spatial domains of our brain activity. Significance: Attention represents a promising strategy to evaluate the quality of the EEG information, and its relevance, in different real-world scenarios. Moreover, it can make it easier to parallelize the computation and, thus, to speed up the analysis of big electrophysiological (e.g., EEG) datasets.
Hateful Memes is a new challenge set for multimodal classification, focusing on detecting hate speech in multimodal memes. Difficult examples are added to the dataset to make it hard to rely on unimodal signals, which means only multimodal models can succeed. According to Kiela,the state-of-the-art methods perform poorly compared to humans (64.73% vs. 84.7% accuracy) on Hateful Memes. I propose a new model that combined multimodal with rules, which achieve the first ranking of accuracy and AUROC of 86.8% and 0.923 respectively. These rules are extracted from training set, and focus on improving the classification accuracy of difficult samples.
We apply persistent homology to the task of discovering and characterizing phase transitions, using lattice spin models from statistical physics for working examples. Persistence images provide a useful representation of the homological data for conducting statistical tasks. To identify the phase transitions, a simple logistic regression on these images is sufficient for the models we consider, and interpretable order parameters are then read from the weights of the regression. Magnetization, frustration and vortex-antivortex structure are identified as relevant features for characterizing phase transitions.
Contrastive representation learning has been recently proved to be very efficient for self-supervised training. These methods have been successfully used to train encoders which perform comparably to supervised training on downstream classification tasks. A few works have started to build a theoretical framework around contrastive learning in which guarantees for its performance can be proven. We provide extensions of these results to training with multiple negative samples and for multiway classification. Furthermore, we provide convergence guarantees for the minimization of the contrastive training error with gradient descent of an overparametrized deep neural encoder, and provide some numerical experiments that complement our theoretical findings
The transition from mild cognitive impairment (MCI) to Alzheimer's disease (AD) is of great interest to clinical researchers. This phenomenon also serves as a valuable data source for quantitative methodological researchers developing new approaches for classification. However, the growth of machine learning (ML) approaches for classification may falsely lead many clinical researchers to underestimate the value of logistic regression (LR), yielding equivalent or superior classification accuracy over other ML methods. Further, in applications with many features that could be used for classifying the transition, clinical researchers are often unaware of the relative value of different selection procedures. In the present study, we sought to investigate the use of automated and theoretically-guided feature selection techniques, and as well as the L-1 norm when applying different classification techniques for predicting conversion from MCI to AD in a highly characterized and studied sample from the Alzheimer's Disease Neuroimaging Initiative (ADNI). We propose an alternative pre-selection technique that utilizes an efficient feature selection based on clinical knowledge of brain regions involved in AD. The present findings demonstrate how similar performance can be achieved using user-guided pre-selection versus algorithmic feature selection techniques. Finally, we compare the performance of a support vector machine (SVM) with that of logistic regression on multi-modal data from ADNI. The present findings show that although SVM and other ML techniques are capable of relatively accurate classification, similar or higher accuracy can often be achieved by LR, mitigating SVM's necessity or value for many clinical researchers.
We propose to apply a 2D CNN architecture to 3D MRI image Alzheimer's disease classification. Training a 3D convolutional neural network (CNN) is time-consuming and computationally expensive. We make use of approximate rank pooling to transform the 3D MRI image volume into a 2D image to use as input to a 2D CNN. We show our proposed CNN model achieves $9.5\%$ better Alzheimer's disease classification accuracy than the baseline 3D models. We also show that our method allows for efficient training, requiring only 20% of the training time compared to 3D CNN models. The code is available online: https://github.com/UkyVision/alzheimer-project.
The COVID-19 has brought about a significant challenge to the whole of humanity, but with a special burden upon the medical community. Clinicians must keep updated continuously about symptoms, diagnoses, and effectiveness of emergent treatments under a never-ending flood of scientific literature. In this context, the role of evidence-based medicine (EBM) for curating the most substantial evidence to support public health and clinical practice turns essential but is being challenged as never before due to the high volume of research articles published and pre-prints posted daily. Artificial Intelligence can have a crucial role in this situation. In this article, we report the results of an applied research project to classify scientific articles to support Epistemonikos, one of the most active foundations worldwide conducting EBM. We test several methods, and the best one, based on the XLNet neural language model, improves the current approach by 93\% on average F1-score, saving valuable time from physicians who volunteer to curate COVID-19 research articles manually.
Respiratory diseases kill million of people each year. Diagnosis of these pathologies is a manual, time-consuming process that has inter and intra-observer variability, delaying diagnosis and treatment. The recent COVID-19 pandemic has demonstrated the need of developing systems to automatize the diagnosis of pneumonia, whilst Convolutional Neural Network (CNNs) have proved to be an excellent option for the automatic classification of medical images. However, given the need of providing a confidence classification in this context it is crucial to quantify the reliability of the model's predictions. In this work, we propose a multi-level ensemble classification system based on a Bayesian Deep Learning approach in order to maximize performance while quantifying the uncertainty of each classification decision. This tool combines the information extracted from different architectures by weighting their results according to the uncertainty of their predictions. Performance of the Bayesian network is evaluated in a real scenario where simultaneously differentiating between four different pathologies: control vs bacterial pneumonia vs viral pneumonia vs COVID-19 pneumonia. A three-level decision tree is employed to divide the 4-class classification into three binary classifications, yielding an accuracy of 98.06% and overcoming the results obtained by recent literature. The reduced preprocessing needed for obtaining this high performance, in addition to the information provided about the reliability of the predictions evidence the applicability of the system to be used as an aid for clinicians.