Drug repositioning is an efficient and promising strategy for traditional drug discovery and development. Many research efforts are focused on utilizing deep-learning approaches based on a heterogeneous network for modeling complex drug–disease associations. Similar to traditional latent factor models, which directly factorize drug–disease associations, they assume the neighbors are independent of each other in the network and thus tend to be ineffective to capture localized information. In this study, we propose a novel neighborhood and neighborhood interaction-based neural collaborative filtering approach (called DRWBNCF) to infer novel potential drugs for diseases. Specifically, we first construct three networks, including the known drug–disease association network, the drug–drug similarity and disease–disease similarity networks (using the nearest neighbors). To take the advantage of localized information in the three networks, we then design an integration component by proposing a new weighted bilinear graph convolution operation to integrate the information of the known drug–disease association, the drug’s and disease’s neighborhood and neighborhood interactions into a unified representation. Lastly, we introduce a prediction component, which utilizes the multi-layer perceptron optimized by the α-balanced focal loss function and graph regularization to model the complex drug–disease associations. Benchmarking comparisons on three datasets verified the effectiveness of DRWBNCF for drug repositioning. Importantly, the unknown drug–disease associations predicted by DRWBNCF were validated against clinical trials and three authoritative databases and we listed several new DRWBNCF-predicted potential drugs for breast cancer (e.g. valrubicin and teniposide) and small cell lung cancer (e.g. valrubicin and cytarabine).

Parameter estimation of photovoltaic (PV) models from experimental current versus voltage (I-V) characteristic curves acts a pivotal part in the modeling a PV system and optimizing its performance. Although many methods have been proposed for solving this PV model parameter estimation problem, it is still challenging to determine highly accurate and reliable solutions. In this paper, this problem is firstly transformed into an optimization problem, and an objective function (OF) is formulated to quantify the overall difference between the experimental and simulated current data. And then, to enhance the performance of original cuckoo search algorithm (CSA), a novel improved cuckoo search algorithm (ImCSA) is proposed, by combining three strategies with CSA. In ImCSA, a quasi-opposition based learning (QOBL) scheme is employed in the population initialization step of CSA. Moreover, a dynamic adaptation strategy is developed and introduced for the step size without Lévy flight step in original CSA. A dynamic adjustment mechanism for the fraction probability (P_a) is proposed to achieve better tradeoff between the exploration and exploitation to increase searching ability. Afterwards, the proposed ImCSA is used for solving the problem of estimating parameters of PV models based on experimental I-V data. Finally, the proposed ImCSA has been demonstrated on the parameter identification of various PV models, i.e., single diode model (SDM), double diode model (DDM) and PV module model (PMM). Experimental results indicate that the proposed ImCSA outperforms the original CSA and its superior performance in comparison with other state-of-the-art algorithms, and they also show that our proposed ImCSA is capable of finding the best values of parameters for the PV models in such effective way for giving the best possible approximation to the experimental I-V data of real PV cells and modules. Therefore, the proposed ImCSA can be considered as a promising alternative to accurately and reliably estimate parameters of PV models.

Background: Artificial intelligence-based disease prediction models have a greater potential to screen COVID-19 patients than conventional methods. However, their application has been restricted because of their underlying black-box nature.Objective: To addressed this issue, an explainable artificial intelligence (XAI) approach was developed to screen patients for COVID-19.Methods: A retrospective study consisting of 1,737 participants (759 COVID-19 patients and 978 controls) admitted to San Raphael Hospital (OSR) from February to May 2020 was used to construct a diagnosis model. Finally, 32 key blood test indices from 1,374 participants were used for screening patients for COVID-19. Four ensemble learning algorithms were used: random forest (RF), adaptive boosting (AdaBoost), gradient boosting decision tree (GBDT), and extreme gradient boosting (XGBoost). Feature importance from the perspective of the clinical domain and visualized interpretations were illustrated by using local interpretable model-agnostic explanations (LIME) plots.Results: The GBDT model [area under the curve (AUC): 86.4%; 95% confidence interval (CI) 0.821–0.907] outperformed the RF model (AUC: 85.7%; 95% CI 0.813–0.902), AdaBoost model (AUC: 85.4%; 95% CI 0.810–0.899), and XGBoost model (AUC: 84.9%; 95% CI 0.803–0.894) in distinguishing patients with COVID-19 from those without. The cumulative feature importance of lactate dehydrogenase, white blood cells, and eosinophil counts was 0.145, 0.130, and 0.128, respectively.Conclusions: Ensemble machining learning (ML) approaches, mainly GBDT and LIME plots, are efficient for screening patients with COVID-19 and might serve as a potential tool in the auxiliary diagnosis of COVID-19. Patients with higher WBC count, higher LDH level, or higher EOT count, were more likely to have COVID-19.

In this paper, we newly propose a holographic ensemble forecasting method (HEFM). First, we use the mutual information and statistical method to select feature variables, which is an ensemble of information about the cross-border multi-source data at the dataset level. Then, we generate multiple training sets by performing diversity sampling with bootstrap, which is an ensemble of information about multiple sample sets at the sampling space level. Next, we construct a multi-model using different artificial intelligence and machine-learning algorithms, which is an ensemble of information about multiple nonlinear heterogeneous models at the forecasting model level. Finally, we use the original features, the forecasting load which is output of the multiple heterogeneous models trained in the first learning, and the actual load of the recent period before each forecasted time to generate a new training set, which is used for the online second learning and final forecasting. This is an ensemble of information about online second learning at the decision level. The ensemble of multi-category multi-state information for four levels (dataset, sampling space, forecasting model, and decision) constitutes the framework of HEFM, whose essence is a forecasting method with comprehensive information integration for the whole life cycle of the forecasting process. We study the load in Guangzhou, China, and New England, USA. Compared to the state-of-the-art forecasting methods, the MAPE of HEFM is reduced by 7.69%-65.77%. The results demonstrate that the forecasting performance may not be improved with the number of algorithms, and that there is a need to understand the positive and negative fusion effect between different algorithms and data characteristics.

Since the variation pattern of load during holidays is different than that of non-holidays, forecasting holiday load is a challenging task. With a focus on this problem, we propose a learning framework based on weighted knowledge transfer for daily peak load forecasting during holidays. First, we select source cities which can provide extra hidden knowledge to improve the forecast accuracy of the load of the target city. Then, all the instances which are from source cities and the target city will be weighted and trained by the improved weighted transfer learning algorithm which is based on the TrAdaBoost algorithm and can decrease negative transfer. We evaluate our method with the classical support vector machine method and a method based on knowledge transfer on a real data set, which includes eleven cities from Guangdong province to illustrate the performance of the method. To solve the problem of limited historical holiday load data, we transfer the data from nearby cities based on the fact that nearby cities in Guangdong province have a similar economic development level and similar load variation pattern. The results of comparative experiments show that the forecasting framework proposed by this paper outperforms these methods in terms of mean absolute percent error and mean absolute scaled error.