Local Attention: Enhancing the Transformer Architecture for Efficient Time Series Forecasting | IEEE Conference Publication | IEEE Xplore