Achieving optimal trade-off for student dropout prediction with multi-objective reinforcement learning.

Student dropout prediction (SDP) in educational research has gained prominence for its role in analyzing student learning behaviors through time series models. Traditional methods often focus singularly on either prediction accuracy or earliness, leading to sub-optimal interventions for at-risk students. This issue underlines the necessity for methods that effectively manage the trade-off between accuracy and earliness. Recognizing the limitations of existing methods, this study introduces a novel approach leveraging multi-objective reinforcement learning (MORL) to optimize the trade-off between prediction accuracy and earliness in SDP tasks. By framing SDP as a partial sequence classification problem, we model it through a multiple-objective Markov decision process (MOMDP), incorporating a vectorized reward function that maintains the distinctiveness of each objective, thereby preventing information loss and enabling more nuanced optimization strategies. Furthermore, we introduce an advanced envelope Q-learning technique to foster a comprehensive exploration of the solution space, aiming to identify Pareto-optimal strategies that accommodate a broader spectrum of preferences. The efficacy of our model has been rigorously validated through comprehensive evaluations on real-world MOOC datasets. These evaluations have demonstrated our model's superiority, outperforming existing methods in achieving optimal trade-off between accuracy and earliness, thus marking a significant advancement in the field of SDP.

A circulating subset of iNKT cells mediates antitumor and antiviral immunity.

Invariant natural killer T (iNKT) cells are a group of innate-like T lymphocytes that recognize lipid antigens. They are supposed to be tissue resident and important for systemic and local immune regulation. To investigate the heterogeneity of iNKT cells, we recharacterized iNKT cells in the thymus and peripheral tissues. iNKT cells in the thymus were divided into three subpopulations by the expression of the natural killer cell receptor CD244 and the chemokine receptor CXCR6 and designated as C0 (CD244-CXCR6-), C1 (CD244-CXCR6+), or C2 (CD244+CXCR6+) iNKT cells. The development and maturation of C2 iNKT cells from C0 iNKT cells strictly depended on IL-15 produced by thymic epithelial cells. C2 iNKT cells expressed high levels of IFN-γ and granzymes and exhibited more NK cell-like features, whereas C1 iNKT cells showed more T cell-like characteristics. C2 iNKT cells were influenced by the microbiome and aging and suppressed the expression of the autoimmune regulator AIRE in the thymus. In peripheral tissues, C2 iNKT cells were circulating that were distinct from conventional tissue-resident C1 iNKT cells. Functionally, C2 iNKT cells protected mice from the tumor metastasis of melanoma cells by enhancing antitumor immunity and promoted antiviral immune responses against influenza virus infection. Furthermore, we identified human CD244+CXCR6+ iNKT cells with high cytotoxic properties as a counterpart of mouse C2 iNKT cells. Thus, this study reveals a circulating subset of iNKT cells with NK cell-like properties distinct from conventional tissue-resident iNKT cells.

A survival analysis based volatility and sparsity modeling network for student dropout prediction.

Student Dropout Prediction (SDP) is pivotal in mitigating withdrawals in Massive Open Online Courses. Previous studies generally modeled the SDP problem as a binary classification task, providing a single prediction outcome. Accordingly, some attempts introduce survival analysis methods to achieve continuous and consistent predictions over time. However, the volatility and sparsity of data always weaken the models' performance. Prevailing solutions rely heavily on data pre-processing independent of predictive models, which are labor-intensive and may contaminate authentic data. This paper proposes a Survival Analysis based Volatility and Sparsity Modeling Network (SAVSNet) to address these issues in an end-to-end deep learning framework. Specifically, SAVSNet smooths the volatile time series by convolution network while preserving the original data information using Long-Short Term Memory Network (LSTM). Furthermore, we propose a Time-Missing-Aware LSTM unit to mitigate the impact of data sparsity by integrating informative missingness patterns into the model. A survival analysis loss function is adopted for parameter estimation, and the model outputs monotonically decreasing survival probabilities. In the experiments, we compare the proposed method with state-of-the-art methods in two real-world MOOC datasets, and the experiment results show the effectiveness of our proposed model.

Synthesis and photocatalytic properties of CuO nanostructures.

CuO nanostructures were grown by decomposition of a mixture of Cu(CH3COO)2 x H2O and NaCl at different temperatures. The nanostructure properties were studied by X-ray diffractometer, scanning electron microscope and Raman spectroscope. Photodegradation activity of the nanostructures towards methyl orange was also examined. CuO spheres and hollow spheres composed of nanoparticles were obtained. CuO nanoparticle size increases with an increase in the growth temperature. More specifically, it increases slowly when the temperature was lower than 280 degrees C and increases dramatically in a higher temperature range. The degradation activity is sensitive to the nanostructure growth temperatures, but the degradation activity varies with the growth temperatures or the size of nanoparticles composing of nanospheres non-monotonously. The hollow spheres composed of nanoparticles grown at 280 degrees C show superior photocatalytic activity towards the degradation of methyl orange than that grown at lower and higher temperatures.