A Hybrid Recommender System Based on Autoencoder and Latent Feature Analysis.

A recommender system (RS) is highly efficient in extracting valuable information from a deluge of big data. The key issue of implementing an RS lies in uncovering users' latent preferences on different items. Latent Feature Analysis (LFA) and deep neural networks (DNNs) are two of the most popular and successful approaches to addressing this issue. However, both the LFA-based and the DNNs-based models have their own distinct advantages and disadvantages. Consequently, relying solely on either the LFA or DNN-based models cannot ensure optimal recommendation performance across diverse real-world application scenarios. To address this issue, this paper proposes a novel hybrid recommendation model that combines Autoencoder and LFA techniques, termed AutoLFA. The main idea of AutoLFA is two-fold: (1) It leverages an Autoencoder and an LFA model separately to construct two distinct recommendation models, each residing in a unique metric representation space with its own set of strengths; and (2) it integrates the Autoencoder and LFA model using a customized self-adaptive weighting strategy, thereby capitalizing on the merits of both approaches. To evaluate the proposed AutoLFA model, extensive experiments on five real recommendation datasets are conducted. The results demonstrate that AutoLFA achieves significantly better recommendation performance than the seven related state-of-the-art models.

Chronic AMD3100 antagonism of SDF-1alpha-CXCR4 exacerbates cardiac dysfunction and remodeling after myocardial infarction.

The role of the SDF-1alpha-CXCR4 axis in response to myocardial infarction is unknown. We addressed it using the CXCR4 antagonist, AMD3100, to block SDF-1alpha interaction with CXCR4 after chronic coronary artery ligation. Chronic AMD3100 treatment decreased ejection fraction and fractional shortening in mice 20days after myocardial infarction compared with vehicle-treated mice (echocardiography). Morphometric analysis showed hearts of AMD3100-treated infarcted mice to have expanded scar, to be hypertrophic (confirmed by myocyte cross-section area) and dilated, with increased LV end systolic and end diastolic dimensions, and to have decreased scar collagen content; p-AKT levels were attenuated and this was accompanied by increased apoptosis. Despite increased injury, c-kit(pos) cardiac progenitor cells (CPCs) were increased in the risk region of AMD3100-treated infarcted mice; CPCs were CD34(neg)/CD45(neg) with the majority undergoing symmetric cell division. c-kit(pos)/MHC(pos) CPCs also increased in the risk region of the AMD3100-treated infarcted group. In this group, GSK-3beta signaling was attenuated compared to vehicle-treated, possibly accounting for increased proliferation and increased cardiac committed MHC(pos) CPCs. Increased proliferation following AMD3100 treatment was supported by increased levels of cyclin D1, a consequence of increased prolyl isomerase, Pin1, and decreased cyclin D1 phosphorylation. In summary, pharmacologic antagonism of CXCR4 demonstrates that SDF-1alpha-CXCR4 signaling plays an important role during and after myocardial infarction and that it exerts pleiotropic salubrious effects, protecting the myocardium from apoptotic cell death, facilitating scar formation, restricting CPC proliferation, and directing CPCs toward a cardiac fate.