Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Mamba ECIS for power marketing customer behavior forecasting using multimodal deep learning.

Weilin Pang1, Xingbo Wang2, Kehuan Li2

  • 1Foshan Power Supply Bureau, Guangdong Power Grid Co., Ltd., Foshan, 528000, China. bangaweng0916@outlook.com.

Scientific Reports
|March 12, 2026
PubMed
Summary

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Precision treatment of COPD based on novel imaging phenotypes: a treatable traits approach.

Respiratory research·2026
Same author

Venocentric perspective on varicocele: summarizing mechanisms and explorations.

Frontiers in physiology·2026
Same author

Early-life triphenyl phosphate (TPhP) exposure impairs the zebrafish dopaminergic system and elevates parkinsonian-like neurotoxicity risk.

Environment international·2026
Same author

[Development of a rapid detection method for a virulent strain of duck enteritis virus based on real-time fluorescence recombinase polymerase amplification].

Sheng wu gong cheng xue bao = Chinese journal of biotechnology·2026
Same author

Co-engineering crystallographic and topological exposures of Zeolitic Imidazolate framework toward superior sulfur redox kinetics in Lithium-sulfur batteries.

Journal of colloid and interface science·2026
Same author

Spin-driven enantioselective regulation of cyclooxygenase-2 activity for rheumatoid arthritis therapy via chiral gold nanohelices.

Nature communications·2026

A new model, Mamba-ECIS, improves energy demand forecasting by integrating multimodal data for better accuracy and interpretability. It captures short-term and long-term trends, outperforming existing methods.

Area of Science:

  • Energy Systems
  • Artificial Intelligence
  • Data Science

Background:

  • Traditional energy forecasting models struggle with complex, multimodal data integration and interpretability.
  • Increasingly complex energy demand and environmental factors necessitate advanced forecasting solutions.

Purpose of the Study:

  • To develop a novel forecasting model, Mamba-ECIS, capable of processing multimodal data simultaneously.
  • To enhance the interpretability of energy demand forecasting models.
  • To improve the accuracy and stability of power demand predictions.

Main Methods:

  • Mamba-ECIS utilizes a dual architecture combined with a causal-consequence attention module.
  • The model incorporates multimodal data to capture both short-term fluctuations and long-term trends.
Keywords:
Causal inferenceDeep learningDemand predictionManbaMultimodal fusionPower marketing

Related Experiment Videos

  • Performance was evaluated on the UCI energy consumption and Pecan Street Dataport datasets.

    • Main Results:

    • Mamba-ECIS demonstrated superior performance compared to benchmark models, with 3-12% improvements across various metrics.
    • Ablation studies confirmed the synergistic effects and individual contributions of model modules.
    • The model achieved more accurate, stable, and interpretable power demand forecasts.

    Conclusions:

    • Mamba-ECIS offers a significant advancement in power demand forecasting.
    • The model's ability to integrate multimodal data and maintain interpretability addresses key limitations of traditional methods.
    • Mamba-ECIS shows high potential for practical application in the energy industry.