Nacionalni portal odprte znanosti

Manually crafted combinatorial features have been the "secret sauce" behind many successful models. For web-scale applications, however, the variety and volume of features make these manually crafted features expensive to create, maintain, and deploy. This paper proposes the Deep Crossing model whic ...

Deep Neural Networks (DNN) have demonstrated superior ability to extract high level embedding vectors from low level features. Despite the success, the serving time is still the bottleneck due to expensive run-time computation of multiple layers of dense matrices. GPGPU, FPGA, or ASIC-based serving ...

Sorption-based atmospheric water harvesting has the potential to realize water production anytime, anywhere, but reaching a hundred-gram high water yield in semi-arid climates is still challenging, although state-of-the-art sorbents have been used. Here, we report a portable and modularized water ha ...