Decoder calibration with ultra small current sample set for intracortical brain–machine interface

The time spent in collecting current samples for decoder calibration and the computational burden brought by high-dimensional neural recordings remain two challenging problems in intracortical brain-machine interfaces (iBMIs). Decoder calibration optimization approaches have been proposed, and neuron selection methods have been used to reduce computational burden. However, few methods can solve both problems simultaneously. In this article, we present a symmetrical-uncertainty-based transfer learning (SUTL) method that combines transfer learning with feature selection. The proposed method uses symmetrical uncertainty to quantitatively measure three indices for feature selection: stationarity, importance and redundancy of the feature. By selecting the stationary features, the disparities between the historical data and current data can be diminished, and the historical data can be effectively used for decoder calibration, thereby reducing the demand for current data. After selecting the important and non-redundant features, only the channels corresponding to them need to work; thus, the computational burden is reduced. The proposed method was tested on neural data recorded from two rhesus macaques to decode the reaching position or grasping gesture. The results showed that the SUTL method diminished the disparities between the historical data and current data, while achieving superior decoding performance with the needs of only ten current samples each category, less than 10% the number of features and 30% the number of neural recording channels. Additionally, unlike most studies on iBMIs, feature selection was implemented instead of neuron selection, and the average decoding accuracy achieved by the former was 6.6% higher.

The development of targeted assays that monitor biomedically relevant proteins is an important step in bridging discovery experiments to large scale clinical studies. Targeted assays are currently unable to scale to hundreds or thousands of targets. We demonstrate the generation of large-scale assays using a novel hybrid nominal mass instrument. The scale of these assays is achievable with the Stellar mass spectrometer through the accommodation of shifting retention times by real-time alignment, while being sensitive and fast enough to handle many concurrent targets. Assays were constructed using precursor information from gas-phase fractionation data-independent acquisition (DIA). We demonstrate the ability to schedule methods from orbitrap and linear ion trap acquired gas-phase fractionation DIA library, and compare the quantification of a matrix-matched calibration curve from orbitrap DIA and linear ion trap parallel reaction monitoring (PRM). Two applications of these proposed workflows are shown with a cerebrospinal fluid neurodegenerative disease protein PRM assay and with a Mag-Net enriched plasma extracellular vesicle protein survey PRM assay. In cerebrospinal fluid, our assay targets proteins discovered previously to be associated with Alzheimer's disease in a small independent sample set. For the Mag-Net enriched plasma survey assay, we observe that proteins selected based on their measurement robustness are still able to capture differences in abundance across disease groups in a small sample set. These highlight the application of highly multiplex, targeted protein assays in clinical research.

The difficulty of feature extraction and the small sample size are two challenges in the field of mechanical fault diagnosis for a long time. Here we propose an intelligent mechanical fault diagnosis method for scenario with small sample datasets. This method can not only diagnose bearing faults but also gear faults, and has strong generalization performance. We use convolutional neural network to realize automatic feature extraction. Through sliding window scanning, one sample set is expanded to three sub-sample sets with different scales to meet the needs of deep learning training. Three convolutional networks are used to extract the features of the subsets respectively to ensure that their useful features are fully extracted. After feature extraction, the feature is reconstructed through feature splicing. Because of the unique advantages of SVM in dealing with small sample sets, we use SVM to classify the reconstructed features. We use the bearing data set collected by Case Western Reserve University in the United States, the bearing fault data set collected by Xi'an Jiaotong University in China, and the gearbox fault data collected by the University of Connecticut in the United States to conduct experiments. The experimental results show that the accuracy of training, validation and testing of the proposed method on the three data sets all reach 100%. This proves that our method can not only tackle the two challenges, but also has high fault diagnosis accuracy and strong generalization performance. It is hoped that our proposed method can contribute to the development of mechanical fault diagnosis.

In this paper a Distribution State Estimator (DSE) tool suitable for real-time monitoring in poorly characterized low voltage networks is presented. An Autoencoder (AE) properly trained with Extreme Learning Machine (ELM) technique is the “brain” of the DSE. The estimation of system state variables, i.e., voltage magnitudes and phase angles is performed with an Evolutionary Particle Swarm Optimization (EPSO) algorithm that makes use of the already trained AE. By taking advantage of historical data and a very limited number of quasi real-time measurements, the presented approach turns possible monitoring networks where information of topology and parameters is not available. Results show improvements in terms of estimation accuracy and time performance when compared to other similar DSE tools that make use of the traditional back-propagation based algorithms for training execution.

The difficulties of obtaining sufficient labeled samples have always been one of the factors hindering deep learning models from obtaining high accuracy in hyperspectral image (HSI) classification. To reduce the dependence of deep learning models on training samples, meta learning methods have been introduced, effectively improving the classification accuracy in small sample set scenarios. However, the existing methods based on meta learning still need to construct a labeled source data set with several pre-collected HSIs, and must utilize a large number of labeled samples for meta-training, which is actually time-consuming and labor-intensive. To solve this problem, this paper proposes a novel unsupervised meta learning method with multiview constraints for HSI small sample set classification. Specifically, the proposed method first builds an unlabeled source data set using unlabeled HSIs. Then, multiple spatial-spectral multiview features of each unlabeled sample are generated to construct tasks for unsupervised meta learning. Finally, the designed residual relation network is used for meta-training and small sample set classification based on the voting strategy. Compared with existing supervised meta learning methods for HSI classification, our method can only utilize HSIs without any label for unsupervised meta learning, which significantly reduces the number of requisite labeled samples in the whole classification process. To verify the effectiveness of the proposed method, extensive experiments are carried out on 8 public HSIs in the cross-domain and in-domain classification scenarios. The statistical results demonstrate that, compared with existing supervised meta learning methods and other advanced classification models, the proposed method can achieve competitive or better classification performance in small sample set scenarios.

Aiming at the poor performance of the modulation recognition method based on deep learning when there are only a few labeled samples in the target sea area, this paper proposes a modulation recognition method for underwater acoustic communication signal with small sample set based on relation network(RN). In this paper, the signal power spectrum is selected as the shallow feature representation of the signal based on the prior information of the signal and the cognition of the task-driven training mode of the relation network. And hence a modulation recognition model based on power spectrum and relation network is designed. This model is optimized by building small sample training tasks in different channels. This training mode improves the ability of the recognition method to quickly classify when only few labeled sample is available in the target sea area. Simulation experiments and the actual signal verify the advanced performance of our algorithm. Compared with existing recognition algorithms, our method requires fewer labeled samples and has better recognition effect.

Regularization is a solution to solve the problem of unstable estimation of covariance matrix with a small sample set in Gaussian classifier. And multi-regularization parameters estimation is more difficult than single parameter estimation. In this paper, KLIM_L covariance matrix estimation is derived theoretically based on MDL (minimum description length) principle for the small sample problem with high dimension. KLIM_L is a generalization of KLIM (Kullback-Leibler information measure) which considers the local difference in each dimension. Under the framework of MDL principle, multi-regularization parameters are selected by the criterion of minimization the KL divergence and estimated simply and directly by point estimation which is approximated by two-order Taylor expansion. It costs less computation time to estimate the multi-regularization parameters in KLIM_L than in RDA (regularized discriminant analysis) and in LOOC (leave-one-out covariance matrix estimate) where cross validation technique is adopted. And higher classification accuracy is achieved by the proposed KLIM_L estimator in experiment.

Aim Niche conservatism is key to understanding species responses to environmental stress such as climate change or arriving in new geographical space such as biological invasion. Halotydeus destructor is an important agricultural pest in Australia and has been the focus of extensive surveys that suggest this species has undergone a niche shift to expand its invasive range inland to hotter and drier environments. We employ modern correlative modelling methods to examine niche conservatism in H. destructor and highlight ecological differences between historical and current distributions.Location Australia and South Africa.Methods We compile comprehensive distribution data sets for H. destructor, representing the native range in South Africa, its invasive range in Australia in the 1960s (40 yr post‐introduction) and its current range in Australia. Using MAXENT, we build correlative models and reciprocally project them between South Africa and Australia and investigate range expansion with models constructed for historical and current data sets. We use several recently developed model exploration tools to examine the climate similarity between native and invasive ranges and subsequently examine climatic variables that limit distributions.Results The invasive niche of H. destructor in Australia transgresses the native niche in South Africa, and the species has expanded in Australia beyond what is predicted from the native distribution. Our models support the notion that H. destructor has undergone a more recent range shift into hotter and drier inland areas of Australia since establishing a stable distribution in the 1960s.Main conclusions Our use of historical and current data highlights that invasion is an ongoing dynamic process and demonstrates that once a species has reached an established range, it may still expand at a later stage. We also show that model exploration tools help understand factors influencing the range of invasive species. The models generate hypotheses about adaptive shifts in H. destructor.

A novel technique, combining the feature extraction mechanisms of a convolutional neural network (CNN) with the classification method of a radial basis probability neural network (RBPNN), is proposed for small sample set modeling and feature knowledge embedding in multi-channel time-varying signal classification. This CNN-RBPNN consists of a signal input layer, signal feature parallel extraction and integration units, and an RBPNN classifier. Each channel signal in a feature extraction unit corresponds to a 1D CNN. The extracted features are represented as feature vectors, and these vectors constitute a comprehensive feature matrix. The RBPNN classifier was designed using signal feature embedding mechanism based on radial basis kernels and the property of combining pattern subclasses into pattern classes to form complex class boundaries. A dynamic clustering algorithm was used to divide each pattern class sample into several subclasses. Typical signal samples in each pattern subclass were designated as kernel centers, in order to achieve signal categories features embedding. This process was also used to determine the number of nodes in the RBPN layer. The RBPN layer outputs were selectively summed in the pattern layer according to kernel center category, which can generate irregular class boundaries, reducing the overlap among different pattern class boundary. The proposed CNN-RBPNN replaces the full-connection layer and classifier unit of conventional CNN with RBPNN, which can extract and represent signals distribution features and structural properties, implement structural and data constraints. This can reduce the structural risks of small sample set modeling. In this study, the properties of CNN-RBPNN are analyzed and an integrated learning algorithm is proposed. An experiment was conducted using 12-lead ECG signals in a seven-classification in the case of small sample set. Results demonstrated that, the correct recognition rate is 5.7% higher than other methods in the experiment, the performance evaluation index also showed significant improvement.

A novel in-depth analysis approach for domain-specific problems based on multidomain data

In this chapter we present a method for learning a compositional model in a minimax entropy framework for modeling object categories with large intra-class variance. The model we learn incorporates the flexibility of a stochastic context free grammar (SCFG) to account for the variation in object structure with the neighborhood constraints of a Markov random field (MRF) to enforce spatial context. We learn the model through a generalized minimax entropy framework that accounts for the dynamic structure of the hierarchical model. We first learn the SCFG parameters using the frequencies of object parts, then pursue spatial relations in order of greatest information gain. The learned model can generalize from a small set of training samples (n < 100) to generate a combinatorially large number of novel instances using stochastic sampling. To verify our learning method and model performance, we present plots of KL divergence minimization as the algorithm proceeds, and show that samples from the model become more realistic as more spatial relations are added. We also show the model accurately predicting missing or undetected parts for top-down recognition along with preliminary results showing that the model can learn a large space of category appearances from a very small (n < 15) number of training samples. This process is similar to “recognition-by-components”, a theory that postulates that biological vision systems recognize objects as composed from a dictionary of commonly appearing 3D structures. Finally, we discuss a compositional boosting algorithm for inference and show examples using it for object recognition. This article is a chapter from the book Object Categorization: Computer and Human Vision Perspectives, edited by Sven Dickinson, Ales Leonardis, Bernt Schiele, and Michael J. Tarr (Cambridge University Press). University of California Los Angeles, Los Angeles, CA. Lotus Hill Research Institute, EZhou, China.

Environmental Risk Assessments (ERA) have been extensively carried out in 2014 by the Company (TOTAL) as part of the Technological Risk Assessment process. In that context it has been applied to nearshore Gabon pipelines and loading terminal located in the vicinity of Mandji Island and Cap Lopez. The approach used is consistent with the methodology already applied to far off-shore installations but in this case was applied for the first time to installations located at a few kilometers from the coast and using a validated set of hindcast current and wind data developed purposely for west coast Africa. The methodology applied takes into account the trajectory of potential oil spills and the probabilities that oil reaches any environmental resources. The tools used are 3D oil drift model (OSCAR) and exposure effect relations coming from the MIRA method developed in Norway. Four oil spill scenarios have been selected as representing most of oil spill cases requiring detailed environmental risk assessment on the basis of the technological risk assessment done on these installations. For each spill scenario, oil drift simulations were conducted in stochastic mode on the basis of 10 years historical wind and current data with 3km resolution. Quantities of oil and associated probabilities of presence were calculated for each cell of the model and comparison could be made with Argos drifter experiments in the case of one scenario. Shoreline habitats, turtles nesting beaches, seabirds, fisheries and mangroves have been selected as the key environmental resources to consider in this study. The results have been used to position the scenarios on the company risk matrix, assess environmental risk and address further needs on risk reduction measures. Results are discussed with some emphasis on the dominating role of current pattern in the area showing that environmental resources at risk are sometimes not the closest to the spill location. Validation aspects of the methodology related to current and Argos drifter experiment data are also discussed.

Healthcare information systems typically collect, store and manage various kinds of data such as illness details, clinical history, essential body parameters, health insurance plans, and other related data towards enabling data processing and analytics to arrive at better decision making with all the clarity and alacrity. To reduce the mortality rate due to heart diseases, it is essential to predict the presence of disease in its budding stage itself. Manual extraction of the useful knowledge from historical data is practically tedious and time-consuming. Machine learning (ML) algorithms are being used to detect and predict something useful out of both historical and current data. Despite the applicability of machine learning algorithms for prediction, the accuracy of prediction is significantly influenced by features used for prediction. Moreover, to meet the needs of evolving data sizes, suitable technologies for data storage also become essential. Based on these two aspects, a comparative analysis has been performed for feature selection using four filter methods, namely, correlation measure, information gain, gain ratio and relief. Further, a scalable architecture using Hadoop framework has been proposed to enable the machine learning algorithms to handle larger datasets while performing prediction task. The impact of the proposed architecture on the performance of machine learning algorithm has been evaluated with benchmark dataset and found to have improved scalability and accuracy.

Questionnaires are widely used in clinical research and mental health studies, but their responses are usually more subjective than clinical biomarkers. Researchers may use latent variable models such as the item response theory (IRT) model to characterize individual ability, but response instability and small sample sizes can lead to increased uncertainty in estimation. Historical data from similar questionnaires may help reduce variation, but they must be incorporated adaptively. The challenge is further compounded by the need to ensure computational efficiency in IRT models with a large number of parameters. In this work, we develop a Bayesian framework for adaptive borrowing in IRT models based on an approximated normalized power prior (NPP) that treats the borrowing weight as a random parameter. The proposed method makes NPP feasible for general IRT models and is coupled with a full Bayesian data augmentation strategy that enables joint estimation of ability and item parameters through an efficient Gibbs sampler. In simulations, the approach adaptively increases borrowing when historical and current data are compatible and automatically downweights conflicting information. Relative to analyzes without borrowing, the method reduces variance and mean squared error while maintaining coverage across a range of test lengths, item discrimination profiles, and historical-current concordance scenarios. We illustrate the method by integrating historical mental health surveys into current assessments, yielding more precise ability estimates with preserved calibration. An efficient implementation is provided in our updated package NPP available on the Comprehensive R Archive Network.

As an important part of emotion research, facial expression recognition is a necessary condition for intelligent interaction between human and machine, which has an important research significance and a potential commercial value. Convolutional neural network (CNN) is an effective method to recognize facial emotions, which can perform feature extraction and classification simultaneously, and can automatically discover multiple levels of representations in data. Due to the fact that there are millions of parameters involved in training the convolutional neural network model, and there is a large demand for marked samples, transfer learning is often used to fine-tune the pre-trained model for a small target sample set. However, there are often some content differences between data sets during deep transfer learning, which will affect the recognition ability of feature extraction. In order to improve the facial expression recognition ability in transfer learning, a hybrid transfer learning model based on an improved convolution restricted boltzmann machine (CRBM) model and a CNN model is proposed in this paper. This method is fused by two learning abilities of these two models. When the pre-trained CNN model is transferred to a small target set, the CRBM is used to replace the full connection layer in the CNN model, and the CRBM layer and the sofmax layer will be retrained on the target set. The added CRBM layer can not only fully connects all feature maps, but can also learn about the unique statistical characteristics about the target set, which eliminates the influence of content differences between data sets, and extracts higher-order statistical features of facial expression images from the target set. The proposed method is evaluated based on four publicly available facial expression databases: JAFFE, FER2013, SFEW and RAF-DB. The new method can achieve better performance than most state-of-the-art methods, and it can effectively prevent the negative influence of transfer learning features between different data sets.