This study proposes MSATE-Net for next-day stock index forecasting. The model combines parallel one-dimensional convolutions with receptive fields of 3, 7, and 15 trading days, a bidirectional LSTM operating entirely inside a historical lookback window, sample-dependent temporal attention, and a residual regularized prediction head. Here, “bidirectional” denotes paired processing of the same observed window; it does not assert time-reversal invariance of financial prices or access to observations after the forecast origin. The globally learned attention temperature controls overall selectivity and is not described as a regime-specific adaptive parameter. Experiments use S&P 500, CSI 300, and Nikkei 225 data; persistence and drift benchmarks; recent forecasting architectures; five-seed uncertainty estimates; expanding-window tests; return and directional metrics; and Diebold–Mariano comparisons. The revised evidence supports lower price-level errors, while directional and significance results are mixed across markets. Because a separate model is fitted in each market, the findings establish cross-market consistency rather than transfer learning.
In recent years, constrained multi-objective optimization problems(CMOPs) remain challenging due to the complex structure of feasible regions, the difficulty of balancing convergence and diversity, and the lack of adaptive operator scheduling mechanisms. To address these issues, this paper proposes a hierarchical reinforcement learning–based subtask-coordinated scheduling method for constrained multi-objective evolutionary algorithm (HRL-SCMOE). The proposed framework employs a two-level architecture, where a high-level agent dynamically schedules subtasks–such as forward-oriented exploration, feasibility-driven exploitation, and diversity guidance–according to the environmental state, while a low-level agent adaptively selects variation operators tailored to each subtask. Both agents are trained using Double Deep Q-Networks (Double DQN) and Prioritized Experience Replay (PER) to enhance stability, sample efficiency, and value estimation reliability. Moreover, the algorithm constructs a set of collaborative information pools targeting different search objectives to maintain balanced exploration between feasible and infeasible regions. An adaptive reward mechanism and soft target updates are also incorporated to improve robustness in hierarchical policy learning. Experimental results on three benchmark test suites and four real-world application domains demonstrate that the proposed method consistently outperforms nine state-of-the-art constrained multi-objective evolutionary algorithms (CMOEAs) in terms of convergence, feasibility, and diversity, thereby confirming its effectiveness and strong general applicability.
Lu-Peng Hao, Yahui Shan, Guang-Yin Jin et al.· Journal of King Saud Univers...· 0 citations
For Industrial Wireless Sensor Networks (IWSNs) serving industrial environmental monitoring tasks, clustering optimization, the core technology for network performance tuning, is a well-recognized NP-hard problem that directly determines the energy efficiency and communication reliability of the entire system. Such IWSNs are typically deployed in large-scale, unattended industrial fields to collect real-time, high-precision environmental data including air quality, water pollution levels and soil parameters. However, inherent constraints like limited node energy supply and unstable wireless links in these scenarios often lead to incomplete data collection and delayed early warning, which directly undermine the reliability of environmental monitoring. To address this challenge, this paper proposes CCNCSO-CRP, a novel energy-efficient clustering routing protocol based on a multi-objective clustering model that jointly considers four key metrics: total network residual energy, average transmission delay, packet loss rate, and the distance from cluster heads to the base station. The protocol is built on the newly designed Chaotic Clonal Niche Cockroach Swarm Optimization (CCNCSO) algorithm, which integrates chaotic initialization and evolutionary strategies including clonal selection and niche preservation to enhance population diversity and convergence speed. Extensive experimental validations are conducted on the CEC2008 and CEC2020 benchmark test suites, where the CCNCSO algorithm outperforms classical meta-heuristic algorithms including Whale Optimization Algorithm (WOA), Osprey Optimization Algorithm (OFA), Komodo Mlipir Algorithm (KMA), Grey Wolf Optimizer (GWO), and Artificial Bee Colony (ABC). Furthermore, experimental evaluations under various IWSN environmental monitoring scenarios show that CCNCSO-CRP outperforms state-of-the-art protocols including LEACH-C, VSSLS-SIACR, and FOAEAUC-SARP by at least 11.5% in extending network lifetime, reduces average delay by no less than 9.5%, and cuts packet loss rate by a minimum of 40.9%. These results validate the effectiveness and superiority of the proposed protocol in improving clustering efficiency and network stability for environmental monitoring IWSNs, and provide reliable technical support for high-performance environmental parameter detection and intelligent early warning systems.
Yunpeng Lv, Tingfa Zhou, Bo Zhou et al.· Scientific Reports· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
