scholarly journals 2P-AGRCFN: Two Phase Attention Gated Recurrent Context Filtering Network for Sequential Recommender Systems

2019 ◽  
Vol 9 (1) ◽  
pp. 3693-3700
Keyword(s):  
Recommender Systems ◽  
Second Phase ◽  
Short Term ◽  
Time Step ◽  
Two Phase ◽  
Current Time ◽  
Sequence Modeling ◽  
Term Dependency ◽  
Hidden States

The recent trends in recommender systems have focused on modeling long-term tastes as well as short-term preferences. The various recurrent architectures have used for sequence modeling in recommender systems, since each state is a combination of current and previous layer output recurrently. Although the Recurrent Neural Networks (RNNs) have the ability for modeling both long-term and short-term dependency to some extent, the monotonic nature of temporal dependency of RNN reduces the effect of short-term interests of the user. Thus final interests of the users can’t be predicted from the hidden states. We propose a Two Phase- Attention Gated Recurrent Context Filtering Network (2P-AGRCF) for dealing with user’s long-term dependency as well as short-term preferences. The first phase of 2P-AGRCFN is performed in the input level by constructing a contextual input using certain number of recent input contexts for handling user’s short-term interests. This can handle the correlation among recent inputs and leads to strong hidden states. In the second phase, the contextual-hidden state is computed by fusing the attention mechanism and the hidden state at current time step, which leads to the effective modeling of overall interest of the user on recommendation. We experiment our model with YooChoose DataSet and it shows efficacy in generating personalized as well as ranked recommendations.

scholarly journals A Novel LSTM Model with Interaction Dual Attention for Radar Echo Extrapolation

2021 ◽  
Vol 13 (2) ◽  
pp. 164
Author(s):  
Chuyao Luo ◽  
Xutao Li ◽  
Yongliang Wen ◽  
Yunming Ye ◽  
Xiaofeng Zhang
Keyword(s):  
Short Term Memory ◽  
Weather Forecast ◽  
Vital Role ◽  
Data Sets ◽  
Short Term ◽  
Learning Techniques ◽  
Radar Echo ◽  
Hidden States ◽  
Better Than

The task of precipitation nowcasting is significant in the operational weather forecast. The radar echo map extrapolation plays a vital role in this task. Recently, deep learning techniques such as Convolutional Recurrent Neural Network (ConvRNN) models have been designed to solve the task. These models, albeit performing much better than conventional optical flow based approaches, suffer from a common problem of underestimating the high echo value parts. The drawback is fatal to precipitation nowcasting, as the parts often lead to heavy rains that may cause natural disasters. In this paper, we propose a novel interaction dual attention long short-term memory (IDA-LSTM) model to address the drawback. In the method, an interaction framework is developed for the ConvRNN unit to fully exploit the short-term context information by constructing a serial of coupled convolutions on the input and hidden states. Moreover, a dual attention mechanism on channels and positions is developed to recall the forgotten information in the long term. Comprehensive experiments have been conducted on CIKM AnalytiCup 2017 data sets, and the results show the effectiveness of the IDA-LSTM in addressing the underestimation drawback. The extrapolation performance of IDA-LSTM is superior to that of the state-of-the-art methods.

Depression in Patients with Schizophrenia in a Psychiatry Hospital in Lagos, Nigeria

2021 ◽  
Vol 6 (1) ◽  
Author(s):  
Akinjola O ◽  
◽  
Lawal RA ◽  
Ojo AM ◽  
Adeosun II ◽  
...  
Keyword(s):  
Social Support ◽  
Depressive Symptoms ◽  
Depressive Disorder ◽  
Second Phase ◽  
Two Phase ◽  
Good Treatment Outcome ◽  
Phase Study ◽  
Poor Social Support ◽  
Schizophrenic Patients

Schizophrenia is a devastating and highly disabling disorder associated with long-term consequences. Treatment is often made difficult by the presence of comorbidities like depression which when considered in management ensures good outcome. This study aimed to determine the prevalence and correlates of depression in schizophrenia. It is a two-phase study involving 320 outpatients recruited by consecutive sampling. The first phase entails confirming diagnosis with Mini International Neuropsychiatric Interview (MINI), psychotic disorder module, assessing socio-demographic characteristic and screening for depressive symptoms with the Beck Depression Inventory (BDI) by a trained assistant. In the second phase, the researcher then assesses for depressive disorder using MINI, depressive disorder module among subjects who screened positive with BDI together with 10% of those who screened negative. Over four-fifth (83.4%) of the participants were less than 50 years, they were mostly females (57.2%), of Yoruba ethnic group (59.7%), Christians (75.6%), and earn below ₦18,000 monthly or nothing (72.2%). Also, a large proportion (86.2%) had good social support. Over a third of the participants were married (38.1%) with about four-fifth of these living with their spouses. The prevalence of depressive symptoms and depressive disorder were 49.7% and 38.4% respectively. Logistic regression revealed that poor social support predicts depressive in Schizophrenia. In conclusion, Depression is common in patients with schizophrenia. Therefore, thorough evaluation of schizophrenic patients is necessary so that, co-morbid depression when present can be detected and considered in management to ensure good treatment outcome.

scholarly journals Temporal Pyramid Recurrent Neural Network

2020 ◽  
Vol 34 (04) ◽  
pp. 5061-5068
Author(s):  
Qianli Ma ◽  
Zhenxi Lin ◽  
Enhuan Chen ◽  
Garrison Cottrell
Keyword(s):  
Large Scale ◽  
Speaker Identification ◽  
Layer Structure ◽  
Input Sequence ◽  
Sequential Data ◽  
Multi Scale ◽  
Sequence Modeling ◽  
Addition Problem ◽  
Hidden States

Learning long-term and multi-scale dependencies in sequential data is a challenging task for recurrent neural networks (RNNs). In this paper, a novel RNN structure called temporal pyramid RNN (TP-RNN) is proposed to achieve these two goals. TP-RNN is a pyramid-like structure and generally has multiple layers. In each layer of the network, there are several sub-pyramids connected by a shortcut path to the output, which can efficiently aggregate historical information from hidden states and provide many gradient feedback short-paths. This avoids back-propagating through many hidden states as in usual RNNs. In particular, in the multi-layer structure of TP-RNN, the input sequence of the higher layer is a large-scale aggregated state sequence produced by the sub-pyramids in the previous layer, instead of the usual sequence of hidden states. In this way, TP-RNN can explicitly learn multi-scale dependencies with multi-scale input sequences of different layers, and shorten the input sequence and gradient feedback paths of each layer. This avoids the vanishing gradient problem in deep RNNs and allows the network to efficiently learn long-term dependencies. We evaluate TP-RNN on several sequence modeling tasks, including the masked addition problem, pixel-by-pixel image classification, signal recognition and speaker identification. Experimental results demonstrate that TP-RNN consistently outperforms existing RNNs for learning long-term and multi-scale dependencies in sequential data.

Effect of PSA-tricom, a pox-viral vaccine in prostate cancer (PCa), on tumor growth rates within 80 days after initiation in nonmetastatic PCa.

2013 ◽  
Vol 31 (6_suppl) ◽  
pp. 57-57 ◽  
Author(s):  
James L. Gulley ◽  
Ravi Amrit Madan ◽  
Wilfred Donald Stein ◽  
Julia Wilkerson ◽  
William L. Dahut ◽  
...  
Keyword(s):  
Prostate Cancer ◽  
Phase Iii ◽  
Clinical Activity ◽  
Short Term ◽  
Two Phase ◽  
Phase Iii Trial ◽  
Castration Resistant ◽  
Viral Vaccine ◽  
Immune Therapies

57 Background: Our understanding of immunotherapies for prostate cancer (PSA-TRICOM, sipuleucel-T, ipilimumab) is incomplete in that such therapies have improved overall survival (OS) without changes in time to progression (TTP) in randomized trials. In an effort to better understand this discrepancy, we evaluated data from studies of PSA-TRICOM. A pox viral vaccine expressing PSA and 3 T-cell co-stimulatory molecules, PSA-TRICOM has demonstrated PSA-specific immune responses and evidence of clinical activity that supported initiation of a currently accruing Phase III trial. An analysis of NCI PCa trials (including a PSA-TRICOM trial) suggests that immune therapies may eventually slow the growth rate (GR) of tumors, leading to unaltered short term TTP, yet improved OS (Stein et al. Clin Can Res. 2011). Methods: PSA-TRICOM was administered to 50 hormone-naïve patients (pts.) with non-metastatic, castration naive PCa in a multi-center trial (ECOG 9802). Pts were treated every 4 weeks for 3 months, then every 12 weeks (preliminary data previously reported, DiPaola, RS et al. ASCO GU 2009). PSA values were used to calculate tumor GR within the first 100 days of treatment. (Pts were given no additional therapies during this time.) As previously described, a two-phase mathematical equation yielded concomitant PSA GR and regression rate constants.(Stein et. al., 2011) Results: See Table. Conclusions: These data suggest that PSA-TRICOM can alter GR significantly within 3 months. If confirmed in future trials, it could explain why vaccines have demonstrated improved OS without improved TTP. A slowing of the GR may not lead to substantial differences in short term TTP, but may enhance OS in the long term. This concept will be evaluated in an international Phase III trial of PSA-TRICOM in minimally symptomatic, metastatic castration-resistant PCa that is currently recruiting pts. Clinical trial information: NCT00108732. [Table: see text]

scholarly journals Virtual interpolation of discrete multi-objective programming solutions with probabilistic operation

2011 ◽  
Vol 22 (4) ◽  
pp. 379-389
Author(s):  
Ricardo C. Silva ◽  
Edilson F. Arruda ◽  
Fabrício O. Ourique
Keyword(s):  
Probability Distributions ◽  
Phase Method ◽  
Second Phase ◽  
Pareto Optimal ◽  
Two Phase ◽  
Term Operation ◽  
Multi Objective ◽  
Multi Objective Programming ◽  
Objective Programming

This work presents a novel framework to address the long term operation of a class of multi-objective programming problems. The proposed approach considers a stochastic operation and evaluates the long term average operating costs/profits. To illustrate the approach, a two-phase method is proposed which solves a prescribed number of K mono-objective problems to identify a set of K points in the Pareto-optimal region. In the second phase, one searches for a set of non-dominated probability distributions that define the probability that the system operates at each point selected in the first phase, at any given operation period. Each probability distribution generates a vector of average long-term objectives and one solves for the Pareto-optimal set with respect to the average objectives. The proposed approach can generate virtual operating points with average objectives that need not have a feasible solution with an equal vector of objectives. A few numerical examples are presented to illustrate the proposed method.

scholarly journals Boosting Memory with a Persistent Memory Mechanism for Remote Sensing Image Captioning

2020 ◽  
Vol 12 (11) ◽  
pp. 1874
Author(s):  
Kun Fu ◽  
Yang Li ◽  
Wenkai Zhang ◽  
Hongfeng Yu ◽  
Xian Sun
Keyword(s):  
Remote Sensing ◽  
Remote Sensing Image ◽  
External Memory ◽  
Image Captioning ◽  
Time Step ◽  
Current Time ◽  
Input Information ◽  
Memory Mechanism ◽  
Persistent Memory

The encoder–decoder framework has been widely used in the remote sensing image captioning task. When we need to extract remote sensing images containing specific characteristics from the described sentences for research, rich sentences can improve the final extraction results. However, the Long Short-Term Memory (LSTM) network used in decoders still loses some information in the picture over time when the generated caption is long. In this paper, we present a new model component named the Persistent Memory Mechanism (PMM), which can expand the information storage capacity of LSTM with an external memory. The external memory is a memory matrix with a predetermined size. It can store all the hidden layer vectors of LSTM before the current time step. Thus, our method can effectively solve the above problem. At each time step, the PMM searches previous information related to the input information at the current time from the external memory. Then the PMM will process the captured long-term information and predict the next word with the current information. In addition, it updates its memory with the input information. This method can pick up the long-term information missed from the LSTM but useful to the caption generation. By applying this method to image captioning, our CIDEr scores on datasets UCM-Captions, Sydney-Captions, and RSICD increased by 3%, 5%, and 7%, respectively.

Community Study of Dementia in the Older Korean Rural Population

2003 ◽  
Vol 37 (5) ◽  
pp. 606-612 ◽  
Author(s):  
Maeng Je Cho ◽  
Jang Kyu Kim ◽  
Guk-Hee Suh
Keyword(s):  
Risk Factors ◽  
Cerebrovascular Disease ◽  
Vascular Dementia ◽  
Rural Population ◽  
Community Study ◽  
Case Group ◽  
Second Phase ◽  
Two Phase ◽  
Korean Elderly

Objective: This study aims to estimate the prevalence of all dementias, including Alzheimer's disease (AD) and vascular dementia (VaD) in a population of Korean elderly and to identify possible risk factors which correlated with specific types of dementia. Method: A two-phase survey, based on a door-to-door survey, was conducted. Initially, the Korean version of the Psychogeriatric Assessment Scale (PAS-K) was administered to all 1037 participants aged 65 years and older. Three hundred and seventy people sampled from the case groups (n = 320) of PAS-K subscales and the non-case group (n = 50) entered the second phase for clinical evaluation. Dementia was defined using the DSM-III-R, NINCDSADRDA and NINDS-AIREN criteria. Results: Among 1037 elderly people aged 65–94 years who completed the interview, 74 cases of dementia were detected, giving an overall age-standardized prevalence (95% CI) of 6.8% (6.1–7.5) (male 6.3%[5.3–7.4]; female 7.1% [6.1–8.0]). The prevalence (95% CI) of AD was 4.2% (3.6–4.7) (male 2.4% [2.0–2.8]; female 5.3% [4.5–6.1]), and it increased with age. The prevalence (95% CI) of VaD was 2.4% (2.0–2.8) (male 3.5% [2.7–4.3]; female 1.6% [1.2–2.1]). Smoking for longer than 30 pack-years significantly increased the risk of VaD (OR = 11.5 [2.8–44,6]). Conclusion: Long-term smoking, much more prevalent in men, may be closely related to higher risk of cerebrovascular disease that leads to vascular dementia.

scholarly journals Deep Learning for Subtyping and Prediction of Diseases: Long-Short Term Memory

2021 ◽  
Author(s):  
Hayrettin Okut
Keyword(s):  
Neural Network ◽  
Short Term Memory ◽  
Short Term ◽  
Time Step ◽  
Term Memory ◽  
The Neural Network ◽  
Long Short Term Memory ◽  
Sequential Information ◽  
Hidden Layer

The long short-term memory neural network (LSTM) is a type of recurrent neural network (RNN). During the training of RNN architecture, sequential information is used and travels through the neural network from input vector to the output neurons, while the error is calculated and propagated back through the network to update the network parameters. Information in these networks incorporates loops into the hidden layer. Loops allow information to flow multi-directionally so that the hidden state signifies past information held at a given time step. Consequently, the output is dependent on the previous predictions which are already known. However, RNNs have limited capacity to bridge more than a certain number of steps. Mainly this is due to the vanishing of gradients which causes the predictions to capture the short-term dependencies as information from earlier steps decays. As more layers in RNN containing activation functions are added, the gradient of the loss function approaches zero. The LSTM neural networks (LSTM-ANNs) enable learning long-term dependencies. LSTM introduces a memory unit and gate mechanism to enable capture of the long dependencies in a sequence. Therefore, LSTM networks can selectively remember or forget information and are capable of learn thousands timesteps by structures called cell states and three gates.

scholarly journals Multimedia Data Modelling Using Multidimensional Recurrent Neural Networks

Symmetry ◽  
2018 ◽  
Vol 10 (9) ◽  
pp. 370
Author(s):  
Zhen He ◽  
Shaobing Gao ◽  
Liang Xiao ◽  
Daxue Liu ◽  
Hangen He
Keyword(s):  
Short Term Memory ◽  
Multimedia Data ◽  
Cross Layer ◽  
Data Modelling ◽  
Short Term ◽  
Time Step ◽  
Machine Learning Approach ◽  
Long Short Term Memory ◽  
Hidden States ◽  
Text Images

Modelling the multimedia data such as text, images, or videos usually involves the analysis, prediction, or reconstruction of them. The recurrent neural network (RNN) is a powerful machine learning approach to modelling these data in a recursive way. As a variant, the long short-term memory (LSTM) extends the RNN with the ability to remember information for longer. Whilst one can increase the capacity of LSTM by widening or adding layers, additional parameters and runtime are usually required, which could make learning harder. We therefore propose a Tensor LSTM where the hidden states are tensorised as multidimensional arrays (tensors) and updated through a cross-layer convolution. As parameters are spatially shared within the tensor, we can efficiently widen the model without extra parameters by increasing the tensorised size; as deep computations of each time step are absorbed by temporal computations of the time series, we can implicitly deepen the model with little extra runtime by delaying the output. We show by experiments that our model is well-suited for various multimedia data modelling tasks, including text generation, text calculation, image classification, and video prediction.

Sign in / Sign up

Export Citation Format

Share Document