Abstract Pancreas segmentation in CT images is fundamental for subsequent diagnosis and qualitative treatment of pancreatic cancer. Since the morphology of the pancreas may be influenced by issues such as class imbalance and boundary blurring across different individuals, segmenting the pancreas from abdominal CT images is a challenging task. To address these issues, this paper proposes a novel pancreas CT image segmentation network, REMC-UNet. First, we introduce the Residual Visual State Space block (ResVSS block) to capture extensive contextual information and effectively extract key features from pancreas CT images. Additionally, we design the Multi-Scale Hybrid Attention (MSHA) to aggregate long-range dependencies and leverage multi-scale spatial information to address the issue of unclear pancreatic boundaries. Finally, we propose the Feature Enhancement block (FE block), which allows the model to focus on global features while also attending to local regions during the feature recovery process. Through experiments on the public NIH dataset, we achieve an average Dice Similarity Coefficient (DSC) of 86.64±4.32%, improving by 2.73% over the baseline model and outperforming other segmentation models.

Switch-UMamba: Dynamic scanning vision Mamba UNet for medical image segmentation.

Deep switching state space model for nonlinear time series forecasting with regime switching

New perspectives on multivariate time series forecasting: Lightweight networks combined with multi-scale hybrid state space models

Hypergraph neural network with state space models for node classification

Skin lesion segmentation network based on state space modeling and convolutional perception

Adaptive multi-stage fusion of hyperspectral and LiDAR data via selective state space models

Understanding the organization of human brain networks has become a central focus in neuroscience, particularly in the study of functional connectivity, which plays a crucial role in diagnosing neurological disorders. Advances in functional magnetic resonance imaging and machine learning techniques have significantly improved brain network analysis. However, traditional machine learning approaches struggle to capture the complex relationships between brain regions, while deep learning methods, particularly Transformer-based models, face computational challenges due to their quadratic complexity in long-sequence modeling. To address these limitations, we propose a Core-Periphery State-Space Model (CP-SSM), an innovative framework for functional connectome classification. Specifically, we introduce Mamba, a selective state-space model with linear complexity, to effectively capture long-range dependencies in functional brain networks. Furthermore, inspired by the core-periphery (CP) organization, a fundamental characteristic of brain networks that enhances efficient information transmission, we design CP-MoE, a CP-guided Mixture-of-Experts that improves the representation learning of brain connectivity patterns. We evaluate CP-SSM on two benchmark fMRI datasets: ABIDE and ADNI. Experimental results demonstrate that CP-SSM surpasses Transformer-based models in classification performance while significantly reducing computational complexity. These findings highlight the effectiveness and efficiency of CP-SSM in modeling brain functional connectivity, offering a promising direction for neuroimaging-based neurological disease diagnosis. Our code is available at https://github.com/m1nhengChen/cpssm.

Lightweight underwater mamba network: A lightweight network with selective state space model for underwater image enhancement

Online transient-based burst localization in water networks using a two-stage ensemble Kalman filter.

Data-driven real-time wind power forecasting based on the dynamic adaptive selective state space model (DA-SSSM)

DGFMamba: Model fine-tuning based on bidirectional state space for domain generalization semantic segmentation

Cross-attention multi-scale state space model for remaining useful life prediction of aircraft engines

State Space Model with dynamic frequency cue for infrared small target detection

The unique demands for analyzing nonlinear time series produced by complex systems have generated a paradigm shift in statistical theory and application in much the same way as nonlinear dynamics have augmented the understanding of specific phenomena in the life and social sciences. Topics covered include: the statistical computation of the fractal dimensions, ergodicity, strategic use of nonlinear model libraries, identifying oscillators, state space analysis and entropy, time delays and the production of emergents, and quantum computing of fractal images. Substantive applications include US unemployment, political affiliation in the Netherlands, bipolar disorder, biomechanics, heart rate complexity, Bitcoin and other market prices, temperature anomalies and climate change, and economic growth.

This study considers a stochastic inventory relocation problem for a one-way, station-based car-sharing system that utilises electric vehicles (EV), where customers' rental demands and rented vehicles' travel distances are uncertain and temporal-spatial imbalance. Workers are hired to relocate vehicles between stations. To maximise the total expected profits that can be collected by the system, a worker must determine whether to relocate an EV, which vehicle to choose, and which station to move the EV to upon their arrival at a rental station. The problem is formulated as a Markov decision process (MDP). A reinforcement learning algorithm is proposed to develop dynamic policies for the problem. The reinforcement learning algorithm uses an approximate value iteration (AVI) algorithm to overcome the computational challenges arising from the extensive state and action space. Action-space restriction and state-space aggregation schemes are developed to enhance the performance of the AVI algorithm. The effectiveness of the proposed modelling and solution methodologies is demonstrated through a comparison of the dynamic policies against benchmark solutions. Additionally, sensitivity analyses are conducted to investigate whether parameter configurations will impact the performance of the dynamic policies.

To address the low trajectory tracking accuracy and limited robustness of conventional reinforcement learning algorithms under complex marine environments involving wind, wave, and current disturbances, this study proposes a proximal policy optimization (PPO) algorithm incorporating an intrinsic curiosity mechanism to solve the unmanned surface vehicle (USV) trajectory tracking control problem. The proposed approach is developed on the basis of a three-degree-of-freedom (3-DOF) USV model and formulated within a Markov decision process (MDP) framework, where a multidimensional state space and a continuous action space are defined, and a multi-objective composite reward function is designed. By incorporating a pure pursuit guidance algorithm, the complexity of engineering implementation is reduced. Furthermore, an improved PPO algorithm integrated with an intrinsic curiosity mechanism is adopted as the trajectory tracking controller, in which the exploration incentives provided by the intrinsic curiosity module (ICM) guide the agent to explore the state space efficiently and converge rapidly to an optimal control policy. The final experimental results indicate that, compared with the conventional PPO algorithm, the improved PPO–ICM controller achieves a reduction of 54.2% in average lateral error and 47.1% in average heading error under simple trajectory conditions. Under the complex trajectory condition, the average lateral error and average heading error are reduced by 91.8% and 41.9%, respectively. These results effectively demonstrate that the proposed PPO–ICM algorithm attains high tracking accuracy and strong generalization capability across different trajectory scenarios, and can provide a valuable reference for the application of intelligent control algorithms in the USV domain.

Hidden Markov models (HMMs) are characterized by an unobservable Markov chain and an observable process—a noisy version of the hidden chain. Decoding the original signal from the noisy observations is one of the main goals in nearly all HMM based data analyses. Existing decoding algorithms such as Viterbi and the pointwise maximum a posteriori (PMAP) algorithm have computational complexity at best linear in the length of the observed sequence, and sub-quadratic in the size of the state space of the hidden chain. We present Quick Adaptive Ternary Segmentation (QATS), a divide-and-conquer procedure with computational complexity polylogarithmic in the length of the sequence, and cubic in the size of the state space, hence, particularly suited for large scale HMMs with relatively few states. It also suggests an effective way of data storage as specific cumulative sums. In essence, the estimated sequence of states sequentially maximizes local likelihood scores among all local paths with at most three segments, and is meanwhile admissible. The maximization is performed only approximately using an adaptive search procedure. Our simulations demonstrate the speedups offered by QATS in comparison to Viterbi and PMAP, along with a precision analysis. An implementation of QATS is in the R-package QATS on GitHub. Supplementary materials for this article are available online.

This study aims to develop an advanced management strategy for the cryptocurrency market This study aims to develop an advanced management strategy for the cryptocurrency market using a deep reinforcement learning framework that integrates real-time sentiment analysis and technical indicators. Beyond improving trading performance, the study frames its approach within a strategic management perspective by emphasizing the alignment of decision-making processes with dynamic market conditions and the proactive handling of financial risks. The core objective is to enhance trading profits while minimizing losses caused by market volatility and emotional bias. Sentiment analysis is conducted using the pre-trained FinRL NLP model to classify market sentiment as positive, negative, or neutral. Historical market data obtained via the Binance API is analyzed in Python, and the models are trained using PPO, A2C, and DQN algorithms. These algorithms incorporate sentiment and technical indicators into the state space. Results show that integrating sentiment analysis improves the effectiveness of decision-making under uncertainty particularly with the A2C algorithm providing more robust performance than traditional strategies. The findings highlight the value of combining sentiment-aware machine learning with strategic risk management to support better-aligned, adaptive investment decisions in volatile environments such as cryptocurrency markets.

Deep Reinforcement Learning (DRL) has become an effective framework of sequential decision-making in high-dimensional and complex control problems, but little has been done to apply it to expressive human movement. This work is a unified DRA algorithm to optimize a dance pose, which an intelligent agent is trained to produce the smooth and stable dance pose and aesthetically compose a pose in a simulated kinematic environment. Generation of dance poses is modeled formally as a Markov Decision Process with incorporation of joint level kinematics, time related dependencies and balance constraints in the state space and pose corrections expressed as a continuous control action. The given framework incorporates pose estimation results and biomechanical constraints to guarantee physical feasibility and motion synthesis that is safe of injuries. Several types of DNR algorithms are tested such as Deep Q-Networks, Proximal Policy Optimization, and Actor-Critic variants to determine their appropriateness to fine-grained pose refinement. The pose accuracy, motion smoothness, energy efficiency and balance stability are well coordinated in a reward function that allows a multi-objective optimization that is in line with technical correctness and artistic quality. Curriculum learning is used to bring a gradual complexity to the poses so that the agent is able to move on to dynamic dance patterns. Substantial experimental investigation shows that policy-gradient-related techniques are more convergent stable and realistic than value-based baselines.