MLP Enhanced CO2 Emission Prediction Model with LWSSA Nature Inspired Optimization

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: Jan. 13, 2025

Abstract Environmental degradation due to the rapid increase in CO₂ emissions is a pressing global challenge, necessitating innovative solutions for accurate prediction and policy development. Machine learning (ML) techniques offer robust approach modeling complex relationships between various factors influencing emissions. Furthermore, ML models can learn interpret significance of each factor’s contribution rise CO 2 . This study proposes novel hybrid framework combining Multi-Layer Perceptron (MLP) with an enhanced Locally Weighted Salp Swarm Algorithm (LWSSA) address limitations traditional optimization techniques, such as premature convergence stagnation locally optimal solutions. The LWSSA improves standard (SSA) by incorporating Mechanism (LWM) Mutation (MM) greater exploration exploitation. LWSSA-MLP achieved accuracy 97% outperformed optimizer-based MLP across several evaluation metrics. A permutation feature analysis identified trade, coal energy, export levels, urbanization, natural resources most influential emissions, offering valuable insights targeted interventions. provides reliable scalable emission prediction, contributing actionable strategies sustainable development environmental resilience.

Language: Английский

Improved salp swarm algorithm based optimization of mobile task offloading

PeerJ Computer Science, Journal Year: 2025, Volume and Issue: 11, P. e2818 - e2818

Published: May 7, 2025

Background The realization of computation-intensive applications such as real-time video processing, virtual/augmented reality, and face recognition becomes possible for mobile devices with the latest advances in communication technologies. This application requires complex computation better user experience decision-making. However, Internet Things (IoT) have computational power limited energy. Executing these computational-intensive tasks on edge may result high energy consumption or latency. In recent times, computing (MEC) has been used modernized offloading this task. MEC, IoT transmit their to servers, which consecutively carry out faster computation. Methods several servers put an upper limit executing concurrent tasks. Furthermore, implementing a smaller size task (1 KB) over server leads improved consumption. Thus, there is need optimum range so that response time will be minimal. evolutionary algorithm best resolving multiobjective Energy, memory, delay reduction together detection achieve. Therefore, study presents salp swarm algorithm-based Mobile Application Offloading Algorithm (ISSA-MAOA) technique MEC. Results harnesses optimization capabilities (ISSA) intelligently allocate between cloud, aiming concurrently minimize consumption, memory usage, reduce completion delays. Through proposed ISSA-MAOA, endeavors contribute enhancement cloud (MCC) frameworks, providing more efficient sustainable solution applications. results research resource management, interactions, enhanced efficiency MCC environments.

Language: Английский