Narrow Results

Functional magnetic resonance imaging (fMRI) has become one of the most important tools for decoding the human brain. Due to the characteristics of the functional MRI samples with high-dimensional features and few samples, there are some difficulties in classifying fMRI data. In this paper, a new algorithm based on transfer learning and the state space model to overcome a few samples is proposed for fMRI classification. First, the fMRI samples need to be preprocessed for data input. Second, the Pre-trainable Vision Transformer (Vit) and Mamba (Previt-Mamba) fusion models are used for feature extraction. Finally, the high-accuracy classification of the fMRI samples is realized by the classifier. The testing results show that Previt-Mamba achieved an average classification accuracy of 80.91% after conducting five-fold cross-validation. At the same time, based on this algorithm model, the mask matrix method is used for reverse research to find human brain region clusters related to human brain change decisions. Through the comparison of classification experiments, it is found that the parietal lobe and frontal lobe of the human brain may have a significant impact on the brain’s decision-making. These research results will help people explore the mysteries of the human brain and reveal the operating mechanism of the human brain.

Electroencephalography (EEG) is a widely used physiological signal to obtain information of brain activity, and its automatic detection holds significant research importance, which saves doctors’ time, improves detection efficiency and accuracy. However, current automatic detection studies face several challenges: large EEG data volumes require substantial time and space for data reading and model training; EEG’s long-term dependencies test the temporal feature extraction capabilities of models; and the dynamic changes in brain activity and the non-Euclidean spatial structure between electrodes complicate the acquisition of spatial information. The proposed method uses range-EEG (rEEG) to extract time-frequency features from EEG to reduce data volume and resource consumption. Additionally, the next-generation state-space model Mamba is utilized as a temporal feature extractor to effectively capture the temporal information in EEG data. To address the limitations of state space models (SSMs) in spatial feature extraction, Mamba is combined with Dynamic Graph Neural Networks, creating an efficient model called DG-Mamba for EEG event detection. Testing on seizure detection and sleep stage classification tasks showed that the proposed method improved training speed by 10 times and reduced memory usage to less than one-seventh of the original data while maintaining superior performance. On the TUSZ dataset, DG-Mamba achieved an AUROC of 0.931 for seizure detection and in the sleep stage classification task, the proposed model surpassed all baselines.

In view of the problem that hydropower station equipment is prone to multiple faults during operation, and to detect SF6 gas leakage faults, this paper proposes a gas detection method based on differential photoacoustic spectroscopy. First, differential absorption photoacoustic spectroscopy technology is used to detect flowing SF6 gas. The system includes a detection module, acquisition module, sampling module, amplification module, control module, and calculation module. Nitrogen is used as power to control the six-way valve for information exchange. SF6 gas is quantitatively injected, samples are transmitted, and the photoacoustic signal is amplified and analyzed. Then, the equipment state space model is constructed, and the particle filter algorithm is applied for state variable estimation. The process is divided into state prediction, update, and resampling. Finally, the residual value is obtained by comparing the real-time measured value and the estimated value, and an adaptive threshold method is added to detect equipment faults and avoid false alarms. The experimental results show that the system in this paper detects a vibration signal of 0.89V when the current is 1000A, and has a good early warning effect on contact system faults and point faults.

Analysis, design and simulation of 126W power supply with better power quality are presented in the proposed work to run an auditorium light emitting diode (LED) light operating at universal AC input mains (90–270V). A single-ended primary inductance converter (SEPIC) topology is designed and driven in continuous conduction mode (CCM) with advance feedback system to maintain constant voltage at output. A proportional integral (PI) controller is also proposed to make the system stable, and stability analysis is discussed in detail with the help of transfer function derived from the state space model. Bode, Nyquist and Polar plots are clearly drawn using the MATLAB tool to claim the system stability. For justification of mathematical analysis, a simulation of the proposed LED driver is also performed in MATLAB–Simulink using sim-power toolbox. The simulation results show the improved value of power quality indices like power factor (PF), total harmonic distortion (THD) and crest factor (CF) with constant rating of 84V, 1.5A at output. Improved PF and reduced THD are under the limit of international standards like IEC-61000-3-2 Class C requirement.

In this work, a zero static error proportional controller for a two-cell DC-DC buck converter is synthesized and analyzed. Under a traditional proportional control scheme, the system presents a constant error of the current supplying the output load. As the proportional feedback gain is increased, the average static error decreases. However, subharmonic oscillations and chaotic behavior emerge beyond successive bifurcations. To achieve zero current and zero voltage static error, we suggest a modification of the traditional proportional controller. By optimizing the feedback gain, the settling time is also decreased. Then, using nonlinear analysis and Lyapunov stability theory, we prove that zero static error is achieved even in the presence of duty cycle saturation. Numerical simulations are presented to confirm our theoretical results.

Modeling and control of dimensional quality is one of deciding factors in current manufacturing competitions, and has always presented a great challenge to both scientists and engineers since for a multi-station machining system, the final product variation is an accumulation from all stations, and the complex non-linear relationship exits between dimensional quality and machining errors. This paper develops a linear state space model using homogeneous transformation to capture the influence of machined errors on dimensional quality, and the explicit expressions for system matrices of the model are explored. The proposed model employs a linear state space form, facilitating the use of the achievements of control theory, information technology and system engineering theory to support engineers supervisory control of physical machining processes, and it also can be used as an analytical engineering tool for efficient and effective faults diagnosis, system plan and design, and optimal sensors allocation. A real machining case illustrates the proposed model.

Combined interaction of all the genes forms a central part of the functional system of a cell. Thus, especially the data-based modeling of the gene expression network is currently one of the main challenges in the field of systems biology. However, the problem is an extremely high-dimensional and complex one, so that normal identification methods are usually not applicable specially if aiming at dynamic models. We propose in this paper a subspace identification approach, which is well suited for high-dimensional system modeling and the presented modified version can also handle the underdetermined case with less data samples than variables (genes). The algorithm is applied to two public stress-response data sets collected from yeast Saccharomyces cerevisiae. The obtained dynamic state space model is tested by comparing the simulation results with the measured data. It is shown that the identified model can relatively well describe the dynamics of the general stress-related changes in the expression of the complete yeast genome. However, it seems inevitable that more precise modeling of the dynamics of the whole genome would require experiments especially designed for systemic modeling.

This study examines the impact of growing political unrest or internal conflict on inbound tourism in the Republic of the Philippines during the period 1994 to 2011. From 2003 onwards, despite formal renunciation of terrorism by one group, the separatist MILF group, acts of violence have continued with increasing political unrest and internal conflict. At the same time an interesting trend of increasing numbers in tourist arrivals was observed from 2003 onwards. The study employs a state space model to test the factors driving tourism during the period 1994 to 2011. The results imply that despite the negative impact of internal conflict and rising inflation, the impact of past income and the inertial effect representing past experience and connectivity to the tourist destination have been sufficient to drive inbound tourism to the Philippines.

This paper proposes a new approach to style analysis of mutual funds in a general state space framework with particle filtering and generalized simulated annealing (GSA). Specifically, we regard the exposure of each style index as a latent state variable in a state space model and employ a Monte Carlo filter as a particle filtering method, where GSA is effectively applied to estimating unknown parameters.

An empirical analysis using data of three Japanese equity mutual funds with six standard style indexes confirms the validity of our method. Moreover, we create fund-specific style indexes to further improve estimation in the analysis.

We propose a statistical strategy to predict differentially regulated genes of case and control samples from time-course gene expression data by leveraging unpredictability of the expression patterns from the underlying regulatory system inferred by a state space model. The proposed method can screen out genes that show different patterns but generated by the same regulations in both samples, since these patterns can be predicted by the same model. Our strategy consists of three steps. Firstly, a gene regulatory system is inferred from the control data by a state space model. Then the obtained model for the underlying regulatory system of the control sample is used to predict the case data. Finally, by assessing the significance of the difference between case and predicted-case time-course data of each gene, we are able to detect the unpredictable genes that are the candidate as the key differences between the regulatory systems of case and control cells. We illustrate the whole process of the strategy by an actual example, where human small airway epithelial cell gene regulatory systems were generated from novel time courses of gene expressions following treatment with(case)/without(control) the drug gefitinib, an inhibitor for the epidermal growth factor receptor tyrosine kinase. Finally, in gefitinib response data we succeeded in finding unpredictable genes that are candidates of the specific targets of gefitinib. We also discussed differences in regulatory systems for the unpredictable genes. The proposed method would be a promising tool for identifying biomarkers and drug target genes.

We propose a state space representation of vector autoregressive model and its sparse learning based on L1 regularization to achieve efficient estimation of dynamic gene networks based on time course microarray data. The proposed method can overcome drawbacks of the vector autoregressive model and state space model; the assumption of equal time interval and lack of separation ability of observation and systems noises in the former method and the assumption of modularity of network structure in the latter method. However, in a simple implementation the proposed model requires the calculation of large inverse matrices in a large number of times during parameter estimation process based on EM algorithm. This limits the applicability of the proposed method to a relatively small gene set. We thus introduce a new calculation technique for EM algorithm that does not require the calculation of inverse matrices. The proposed method is applied to time course microarray data of lung cells treated by stimulating EGF receptors and dosing an anticancer drug, Gefitinib. By comparing the estimated network with the control network estimated using non-treated lung cells, perturbed genes by the anticancer drug could be found, whose up- and down-stream genes in the estimated networks may be related to side effects of the anticancer drug.

By surveying recent studies by molecular biologists and cancer geneticists, in this chapter we have proposed general stochastic models of car-cinogenesis and provided biological evidences for these models. Because most of these models are quite complicated far beyond the scope of the MVK two-stage model, the traditional Markov theory approach becomes too complicated to obtain analytical results. To develop these stochastic models, in this chapter we thus propose an alternative approach through stochastic differential equations. Given observed cancer incidence data, we further combine these stochastic models with statistical models to develop state space models for carcinogenesis. By using these state space models, we then develop a generalized Bayesian procedure to estimate the unknown parameters and to predict state variables via multi-level Gibbs sampling procedures. In this chapter we have used the multi-event model as an example to illustrate our modeling approach and some basic theories.

In this chapter we survey recent studies by molecular biologists and cancer geneticists and propose some general stochastic models of carcinogenesis. To develop analytic results for these stochastic models, because the traditional Markov theory approach becomes too complicated to be of much use, in this chapter we propose an alternative approach through stochastic equations. Given observed cancer incidence data, we further combine these stochastic models with statistical models to develop state space models for carcinogenesis. By using these state space models, we then develop a generalized Bayesian method and a predicted inference procedure to estimate the unknown parameters and to predict state variables via multi-level Gibbs sampling procedures. In this chapter we use the extended multi-event model and mixture model as examples to illustrate our modeling approach and some basic theories.

Based on recent biological studies, in this chapter we have developed a stochastic model for human colon cancer involving five different pathways. These pathways are: the sporadic LOH pathway (about 70–75%), the familial LOH pathway (about 10–15%), the FAP pathway (about 1%), the sporadic MSI pathway (about 10%) and the HNPCC pathway (about 5%). For this model, we have combined the data augmentation method (equivalent to the EM algorithm in sampling theory framework) with the genetic algorithm (GA algorithm) and the state space model to estimate the genetic parameters of these pathways. We use the Bayesian approach to estimate the parameters through the posterior modes of the parameters by combining the genetic algorithm with the mean numbers of state variables. We have applied this model to fit and analyze the SEER data of human colon cancers from NCI/NIH. Our results indicate that the model not only provides a logical avenue to incorporate biological information but also fits the data much better than other models including the four-stage single pathway model. This model not only would provide more insights into human colon cancer but also would provide useful guidance for its prevention and control and for prediction of future cancer cases.

This article illustrates how to use stochastic models and state space models to assess risk of environmental agents. In this state space model, the stochastic system model is the general stochastic multi-stage model of carcinogenesis whereas the observation model is a statistical model based on cancer incidence data. To analyze the stochastic system model, in this chapter we introduce the stochastic equations for the state variables and derive the probability distributions of these variables. In this chapter we also introduce the genetic algorithm, the multi-level Gibbs sampling procedure and the predicted inference procedure to estimate the unknown genetic parameters and to predict the state variables. As an application, the model and the method are used to illustrate how the arsenic in drinking water induces the bladder cancer in human beings using cancer incidence data given in Morale et al. (2000). Our analysis clearly indicated that in the induction of human bladder cancer the arsenic in drinking water was both an initiator and a weak promoter. These results are not possible by purely statistical methods.