Attention-Based Models for Multivariate Time Series Forecasting: Multi-step Solar Irradiation Prediction

Bangladesh's subtropical climate with an abundance of sunlight throughout the greater portion of the year results in increased effectiveness of solar panels. Solar irradiance forecasting is an essential aspect of grid-connected photovoltaic systems to efficiently manage solar power's variation and uncertainty and to assist in balancing power supply and demand. This is why it is essential to forecast solar irradiation accurately. Many meteorological factors influence solar irradiation, which has a high degree of fluctuation and uncertainty. Predicting solar irradiance multiple steps ahead makes it difficult for forecasting models to capture long-term sequential relationships. Attention-based models are widely used in the field of Natural Language Processing for their ability to learn long-term dependencies within sequential data. In this paper, our aim is to present an attention-based model framework for multivariate time series forecasting. Using data from two different locations in Bangladesh with a resolution of 30 min, the Attention-based encoder-decoder, Transformer, and Temporal Fusion Transformer (TFT) models are trained and tested to predict over 24 steps ahead and compared with other forecasting models. According to our findings, adding the attention mechanism significantly increased prediction accuracy and TFT has shown to be more precise than the rest of the algorithms in terms of accuracy and robustness. The obtained mean square error (MSE), the mean absolute error (MAE), and the coefficient of determination (R2) values for TFT are 0.151, 0.212, and 0.815, respectively. In comparison to the benchmark and sequential models (including the Naive, MLP, and Encoder-Decoder models), TFT has a reduction in the MSE and MAE of 8.4-47.9% and 6.1-22.3%, respectively, while R2 is raised by 2.13-26.16%. The ability to incorporate long-distance dependency increases the predictive power of attention models.

Ultrasound classification of breast masses using a comprehensive Nakagami imaging and machine learning framework

In this study we investigate the potential of parametric images formed from ultrasound B-mode scans using the Nakagami distribution for non-invasive classification of breast lesions and characterization of breast tissue. Through a sliding window technique, we generated seven types of Nakagami images for each patient scan in our dataset using basic and as well as derived parameters of the Nakagami distribution. To determine the suitable window size for image generation, we conducted an empirical analysis using 4 windows, which includes 3 column windows of lengths 0.1875 mm, 0.45 mm and 0.75 mm and widths of 0.002 mm, along with the standard square window with sides equal to three times the pulse length of incident ultrasound. From the parametric image sets generated using each window, we extracted a total of 72 features that consisted of morphometric, elemental and hybrid features. To our knowledge no other literature has conducted such a comprehensive analysis of Nakagami parametric images for the classification of breast lesions. Feature selection was performed to find the most useful subset of features from each of the parametric image sets for the classification of breast cancer. Analyzing the classification accuracy and Area under the Receiver Operating Characteristic (ROC) Curve (AUC) of the selected feature subsets, we determined that the selected features acquired from Nakagami parametric images generated using a column window of length 0.75 mm provides the best results for characterization of breast lesions. This optimal feature set provided a classification accuracy of 93.08%, an AUC of 0.9712, a False Negative Rate (FNR) of 0%, and a very low False Positive Rate (FPR) of 8.65%. Our results indicate that the high accuracy of such a procedure may assist in the diagnosis of breast cancer by helping to reduce false positive diagnoses.