Increasingly, more people are suffering from the effects of air pollution. This study took Beijing as an example and proposed an attention-based air quality predictor (AAQP) that could better protect people from air pollution. The AAQP is a seq2seq model, and it exploits historical air quality data and weather data to predict future air quality indexes. Although existing research has promoted seq2seq for air quality prediction, there are still two problems. First, the seq2seq has a slow training speed so the original RNN in the encoder was replaced with a fully connected encoder to accelerate the training process. Position embedding was also introduced to help the fully connected encoder find the sequential relationships among source sequences. Another problem is error accumulation caused by recurrent prediction. Accordingly, the n-step recurrent prediction was proposed to solve this problem. The experimental results validated that the AAQP with n-step recurrent prediction had better performance than the related arts since the error accumulation was reduced, and the training time was significantly decreased compared with the original seq2seq attention model.
Low-rank tensor completion methods have been advanced recently for modeling sparsely observed data with a multimode structure. However, low-rank priors may fail to interpret the model factors of general tensor objects. The most common method to address this drawback is to use regularizations together with the low-rank priors. However, due to the complex nature and diverse characteristics of real-world multiway data, the use of a single or a few regularizations remains far from efficient, and there are limited systematic experimental reports on the advantages of these regularizations for tensor completion. To fill these gaps, we propose a modified CP tensor factorization framework that fuses the l₂ norm constraint, sparseness (l₁ norm), manifold, and smooth information simultaneously. The factorization problem is addressed through a combination of Nesterov's optimal gradient descent method and block coordinate descent. Here, we construct a smooth approximation to the $l_1$ norm and TV norm regularizations, and then, the tensor factor is updated using the projected gradient method, where the step size is determined by the Lipschitz constant. Extensive experiments on simulation data, visual data completion, intelligent transportation systems, and GPS data of user involvement are conducted, and the efficiency of our method is confirmed by the results. Moreover, the obtained results reveal the characteristics of these commonly used regularizations for tensor completion in a certain sense and give experimental guidance concerning how to use them.
No abstract
Bayesian Personalized Ranking (BPR) is a representative pairwise learning method for optimizing recommendation models. It is widely known that the performance of BPR depends largely on the quality of negative sampler. In this paper, we make two contributions with respect to BPR. First, we find that sampling negative items from the whole space is unnecessary and may even degrade the performance. Second, focusing on the purchase feedback of E-commerce, we propose an effective sampler for BPR by leveraging the additional view data. In our proposed sampler, users' viewed interactions are considered as an intermediate feedback between those purchased and unobserved interactions. The pairwise rankings of user preference among these three types of interactions are jointly learned, and a user-oriented weighting strategy is considered during learning process, which is more effective and flexible. Compared to the vanilla BPR that applies a uniform sampler on all candidates, our view-enhanced sampler enhances BPR with a relative improvement over 37.03% and 16.40% on two real-world datasets. Our study demonstrates the importance of considering users' additional feedback when modeling their preference on different items, which avoids sampling negative items indiscriminately and inefficiently.
A 10-week study was conducted to investigate the effects of feeding rate and frequency on growth performance, digestion and nutrients balances of Atlantic salmon (Salmo salar) in replicated recirculating aquaculture systems (RAS). Replicated groups of juvenile salmon weighing 90 AE 2.5 g (mean AE SD) were fed a commercial feed (21.63 MJ kg À1 gross energy) to designed feeding rate (1.4%, 1.6% and 1.8% body weight day À1 , BW day À1 ) and feeding frequency (2 and 4 meals day À1 ) combinations. Specific growth ratio varied between 1.15 AE 0.02 and 1.37 AE 0.16% day À1 , and feed conversion ratio ranged from 0.96 AE 0.03 to 1.16 AE 0.02. The nitrogen and phosphorus retention rates were from 36.50 AE 1.94 to 47.08 AE 5.23% and from 20.42 AE 1.05 to 38.59 AE 2.80%. Apparent digestibility coefficients (ADC) in dry matter, protein, lipid and energy showed no significant differences for all groups. However, fish fed at 1.6% BW day À1 and 4 meal day À1 groups had relatively better growth and nutrient retention efficiency compared to other groups. In addition, concentrations of nitrogenous and phosphorous compounds were also detected in this study. These results suggested that salmon of 100-200 g in RAS could in practice be fed at 1.6% BW day À1 and 4 meals day À1 , taking environmental impacts into account.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.