Type 2 diabetes mellitus (T2DM) has been identified as one of the most challenging chronic diseases to manage. In recent years, the incidence of T2DM has increased, which has seriously endangered people's health and life quality. Glycosylated hemoglobin (HbA1c) is the gold standard clinical indicator of the progression of T2DM. An accurate prediction of HbA1c levels not only helps medical workers improve the accuracy of clinical decision-making but also helps patients to better understand the clinical progression of T2DM and conduct self-management to achieve the goal of controlling the progression of T2DM. Therefore, we introduced the long short-term memory (LSTM) neural network to predict patients' HbA1c levels using time sequential data from electronic medical records (EMRs). We added the self-attention mechanism based on the traditional LSTM to capture the long-term interdependence of feature elements and which ensure that the memory was more profound and effective, and used the gradient search technology to minimize the mean square error of the predicted value of the network and the real value. LSTM with the self-attention mechanism performed better than the traditional deep learning sequence prediction method. Our research provides a good reference for the application of deep learning in the field of medical health management.
The future of smart grids promises unprecedented flexibility in energy management. Therefore, it is crucial to accurately predict the energy load demand of individual power grid stations and overall levels. In this paper, a Sequence to Sequence learning method based on long short-term memory neural network (Seq2Seq) forecast model is proposed to predict the load of the community microgrid. In particular, this paper studies the effectiveness of Seq2Seq in community microgrid load forecasting, avoiding overfitting through weight attenuation regularization, and reducing the impact of interference and noise on training data. The proposed method is implemented on the base load data set of multi-user. The results of Seq2Seq were compared with the results of two well-known methods, one is traditional LSTM and the other is Support Vector Machine (SVM) on the same datasets. Experimental results show that compared with SVM and traditional LSTM, Seq2Seq has better learning effect. Finally, we calculate the performance index. When using load data to train and test the model, it was noted that compared with traditional LSTM and SVM, the root mean square error (RMSE) of Seq2Seq was decreased by 44.07% and 64.06%, respectively. Our research offers a decent reference for the application of deep learning in the field of energy management.
An open question that has existed for some time now is how to preserve rank in the AHP when a new alternative is added or when one is deleted. The essential conditions are that all judgments be consistent and all elements are independent; these have not been fully considered by the AHP critics and defenders. When a new alternative is added or when one is deleted, rank should be preserved when the conditions are satisfied. The weighted geometric mean aggregation rule is proposed to achieve the desired outcome. A proof demonstrates that the weighted geometric mean aggregation rule can preserve rank in the normalized priority vector. Finally, the causes of rank reversal are analyzed: the principal eigenvector approach and the relative mode, and derive that they are not the real reasons of rank reversal.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.