No abstract

No abstract

No abstract

Lead halide perovskites have shown great potential in photonic synaptic devices owing to their notable optical properties. However, the inherent toxicity of lead and complex configuration of the device hinder the further development of perovskites‐based photonic synaptic devices. Here, lead‐free double perovskite Cs2AgSbBr6 is first used to fabricate photonic synaptic devices. The photonic synaptic device with a simple structure exhibits decent reproducibility and stability. By tuning external light signals, photonic synaptic devices based on Cs2AgSbBr6 can mimic primary synaptic behaviors, such as excitatory postsynaptic current, paired pulse facilitation, short‐term memory (STM), long‐term memory (LTM), the transfer between STM and LTM, and the “learning‐forgetting‐relearning” process. Furthermore, the temperature effect on the learning and memory process is investigated. Importantly, logic functions of “OR” and “AND” operations are realized. This work suggests the great potential of Cs2AgSbBr6 in future neuromorphic computation.

Convolutional Neural Network (CNN) achieved satisfying performance in click-through rate (CTR) prediction in recent studies. Since features used in CTR prediction have no meaningful sequence in nature, the features can be arranged in any order. As CNN learns the local information of a sample, the feature sequence may influence its performance significantly. However, this problem has not been fully investigated. This paper firstly investigates whether and how the feature sequence affects the performance of the CNN-based CTR prediction method. As the data distribution of CTR prediction changes with time, the best current sequence may not be suitable for future data. Two multi-sequence models are proposed to learn the information provided by different sequences. The first model learns all sequences using a single feature learning module, while each sequence is learnt individually by a feature learning module in the second one. Moreover, a method of generating a set of embedding sequences which aims to consider the combined influence of all feature pairs on feature learning is also introduced. The experiments are conducted to demonstrate the effectiveness and stability of our proposed models in the offline and online environment on both the benchmark Avazu dataset and a real commercial dataset.