Reduce the Cold Start of COVID-19 In-hospital Mortality Prediction Models via Transfer Learning

At the beginning of the breakout of a new disease, the healthcare community almost always has little experience in treating patients of this kind. Similarly, due to insufficient patient records at the early stage of a pandemic, it is difficult to train an in-hospital mortality prediction model specific to the new disease. We call this the 'cold start' problem of mortality prediction models. In this paper, we aim to study the cold start problem of 3-days ahead COVID-19 mortality prediction models by the following two steps: (i) Train XGBoost [1] and logistic regression 3-days ahead mortality prediction models on MIMIC3, a publicly available ICU patient dataset [2];(ii) Apply those MIMIC3 models to COVID-19 patients and then use the prediction scores as a new feature to train COVID-19 3-days ahead mortality prediction models. Retrospective experiments are conducted on a real-world COVID-19 patient dataset(n = 1,287) collected in US from June 2020 to February 2021 with a mixed cohort of both ICU and Non-ICU patients. Since the dataset is imbalanced(death rate = 7.8%), we primarily focus on the relative improvement of AUPR. We trained models with and without MIMIC3 scores on the first 200, 400,..., 1000 patients respectively and then tested on the next 200 incoming patients. The results show a diminishing positive transfer effect of AUPR from 5.36% for the first 200 patients(death rate = 5.5%) to 3.58% for all 1,287 patients. Meanwhile the AUROC scores largely remain unchanged, regardless of the number of patients in the training set. What's more, the p-value of t-test suggests that the cold start problem disappears for a dataset larger than 600 COVID-19 patients. To conclude, we demonstrate the possibility of mitigating the cold start problem via the proposed method. © 2022 IEEE.

A Hybrid Recommender System for Improving Rating Prediction of Movie Recommendation

Because of COVID-19 pandemic, online movies are now extremely popular. While the movie theaters have not serviced and people are staying quarantine, movies are the best choice for relaxing and treating stress. In present, recommender systems are widely integrated into many platforms of movie applications. A hybrid recommender system is one promising technique to improve the system performance, especially for cold-start, data sparsity, and scalability. This paper proposed a hybrid of matrix factorization, biased matrix factorization, and factor wise matrix factorization to solve all mentioned drawback problems. Simulation shows that the proposed hybrid algorithm can decrease approximately 11.91% and 10.70% for RMSE and MAE, respectively, when compared with the traditional methods. In addition, the proposed algorithm is capable of scalability. While the number of datasets is tremendously increased by 10 times, it is still effectively executed. © 2022 IEEE.

Prediction-driven collaborative emergency medical resource allocation with deep learning and optimization

This study addresses two key issues, ie, the “cold-start problem” in transmission prediction of new or rare epidemics and the collaborative allocation of emergency medical resources considering multiple objectives. These two issues have not yet been well addressed in data-driven emergency medical resource allocation systems. A decision support prediction-then-optimization framework combing deep learning and optimization is developed to address these two issues. Two transfer learning based convolutional neural network models are built for epidemic transmission predictions in the initial and the subsequent outbreak regions using transfer learning to deal with the “cold-start problem”. A prediction-driven collaborative emergency medical resource allocation model is built to address the issue of collaborative decisions by simultaneously considering the inter- and intra-echelon resource flows in a multi-echelon system and considering the efficiency and fairness as the objective functions. A case study of the COVID-19 pandemic shows that combining transfer learning and convolutional neural networks can improve the performances of epidemic transmission predictions, and good predictions can improve both the efficiency and fairness of emergency medical resource allocation decisions. Moreover, the computational results show that the prediction errors are asymmetrically amplified in the optimization stage, and the shortage of the resource reserve quantity mediates the asymmetrical amplification effect. © Operational Research Society 2022.