This paper addresses the challenge of multivariate time series forecasting, which involves capturing both short-term and long-term patterns in real-world applications such as solar energy output, electricity consumption, and traffic jam situations. Traditional methods like autoregressive models and Gaussian Processes often fail to handle these complex patterns. To tackle this issue, the authors propose a novel deep learning framework called Long- and Short-term Time-series Network (LSTNet). LSTNet combines Convolutional Neural Networks (CNN) and Recurrent Neural Networks (RNN) to extract short-term local dependencies and discover long-term trends. Additionally, it incorporates a traditional autoregressive model to address scale insensitivity issues. The evaluation on real-world datasets with complex patterns shows that LSTNet outperforms several state-of-the-art methods, demonstrating its effectiveness in capturing both short-term and long-term patterns. The paper also includes an ablation study to validate the importance of each component in the LSTNet architecture and discusses future directions for improvement.This paper addresses the challenge of multivariate time series forecasting, which involves capturing both short-term and long-term patterns in real-world applications such as solar energy output, electricity consumption, and traffic jam situations. Traditional methods like autoregressive models and Gaussian Processes often fail to handle these complex patterns. To tackle this issue, the authors propose a novel deep learning framework called Long- and Short-term Time-series Network (LSTNet). LSTNet combines Convolutional Neural Networks (CNN) and Recurrent Neural Networks (RNN) to extract short-term local dependencies and discover long-term trends. Additionally, it incorporates a traditional autoregressive model to address scale insensitivity issues. The evaluation on real-world datasets with complex patterns shows that LSTNet outperforms several state-of-the-art methods, demonstrating its effectiveness in capturing both short-term and long-term patterns. The paper also includes an ablation study to validate the importance of each component in the LSTNet architecture and discusses future directions for improvement.