Capturing Feature-Level Irregularity in Disease Progression Modeling

Zheng, Kaiping; Wang, Wei; Gao, Jinyang; Ngiam, Kee Yuan; Ooi, Beng Chin; Yip, Wei Luen James

doi:10.1145/3132847.3132944

Cited by 25 publications

(13 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…With the development of deep learning, neural networks have been widely used in many domains, such as disease diagnosis [42], translation [1], and recommendation [8]. Song et al [25] propose a novel deep neural network-based architecture that models the combination of long-term static and short-term temporal user preferences to improve the recommendation performance.…”

Section: Neural Network-based Recommendationmentioning

confidence: 99%

Group-Based Recurrent Neural Networks for POI Recommendation

Zheng

et al. 2020

ACM/IMS Trans. Data Sci.

View full text Add to dashboard Cite

With the development of mobile Internet, many location-based services have accumulated a large amount of data that can be used for point-of-interest (POI) recommendation. However, there are still challenges in developing an unified framework to incorporate multiple factors associated with both POIs and users due to the heterogeneity and implicity of this information. To alleviate the problem, this work proposes a novel group-based method for POI recommendation jointly considering the reviews, categories, and geographical locations, called the Group-based Temporal Sentiment-Aspect-Region Recurrent Neural Network (GTSAR-RNN). We divide the users into different groups and then train an individual RNN for each group with the goal of improving its pertinence. In GTSAR-RNN, we consider not only the effects of temporal and geographical contexts but also the users' sentimental opinions on locations. Experimental results show that GTSAR-RNN acquires significant improvements over the baseline methods on real datasets.

show abstract

Section: Neural Network-based Recommendationmentioning

confidence: 99%

Group-Based Recurrent Neural Networks for POI Recommendation

Zheng

et al. 2020

ACM/IMS Trans. Data Sci.

View full text Add to dashboard Cite

show abstract

“…Once developed, chronic diseases rarely disappear and usually have a longer lasting impact on future visits than acute diseases. When each input vector includes one patient visit’s information, Bai et al [46,62] improved LSTM prediction accuracy by learning different time decay factors for differing diseases to reflect this. We can make this more explicit to help LSTM remember long- span history and further boost prediction accuracy.…”

Section: Semi-automatically Extracting Predictive and Clinically Meanmentioning

confidence: 99%

A roadmap for semi-automatically extracting predictive and clinically meaningful temporal features from medical data for predictive modeling

Luo

2019

Global Transitions

View full text Add to dashboard Cite

Predictive modeling based on machine learning with medical data has great potential to improve healthcare and reduce costs. However, two hurdles, among others, impede its widespread adoption in hdealthcare. First, medical data are by nature longitudinal. Pre-processing them, particularly for feature engineering, is labor intensive and often takes 50–80% of the model building effort. Predictive temporal features are the basis of building accurate models, but are difficult to identify. This is problematic. Healthcare systems have limited resources for model building, while inaccurate models produce sub-optimal outcomes and are often useless. Second, most machine learning models provide no explanation of their prediction results. However, offering such explanations is essential for a model to be used in usual clinical practice. To address these two hurdles, this paper outlines: 1) a data-driven method for semi-automatically extracting predictive and clinically meaningful temporal features from medical data for predictive modeling; and 2) a method of using these features to automatically explain machine learning prediction results and suggest tailored interventions. This provides a roadmap for future research.

show abstract

“…Some related works introduce the time information or the interval information of the events into the model to solve the problem of inconsistent sampling frequency [9][10][11][12]. For example, Che et al multiply the hidden state by a time decay factor before calculating the next hidden state in Gated Recurrent Unit (GRU) [10].…”

Section: B Deep Sequential Models For Ehrmentioning

confidence: 99%

“…For example, Che et al multiply the hidden state by a time decay factor before calculating the next hidden state in Gated Recurrent Unit (GRU) [10]. Zheng et al balance the inheritance and update of hidden states based on the time decay function when updating the hidden layer state of GRU [11]. Bai et al propose the Timeline model to model the decay rate of different events affecting patients [9].…”

Section: B Deep Sequential Models For Ehrmentioning

confidence: 99%

Predictive Multi-level Patient Representations from Electronic Health Records

Wang

Liu

et al. 2019

2019 IEEE International Conference on Bioinformatics and Biomedicine (BIBM)

View full text Add to dashboard Cite

The advent of the Internet era has led to an explosive growth in the Electronic Health Records (EHR) in the past decades. The EHR data can be regarded as a collection of clinical events, including laboratory results, medication records, physiological indicators, etc, which can be used for clinical outcome prediction tasks to support constructions of intelligent health systems. Learning patient representation from these clinical events for the clinical outcome prediction is an important but challenging step. Most related studies transform EHR data of a patient into a sequence of clinical events in temporal order and then use sequential models to learn patient representations for outcome prediction. However, clinical event sequence contains thousands of event types and temporal dependencies. We further make an observation that clinical events occurring in a short period are not constrained by any temporal order but events in a long term are influenced by temporal dependencies. The multi-scale temporal property makes it difficult for traditional sequential models to capture the short-term co-occurrence and the long-term temporal dependencies in clinical event sequences. In response to the above challenges, this paper proposes a Multilevel Representation Model (MRM). MRM first uses a sparse attention mechanism to model the short-term co-occurrence, then uses interval-based event pooling to remove redundant information and reduce sequence length and finally predicts clinical outcomes through Long Short-Term Memory (LSTM). Experiments on real-world datasets indicate that our proposed model largely improves the performance of clinical outcome prediction tasks using EHR data.

show abstract

Capturing Feature-Level Irregularity in Disease Progression Modeling

Cited by 25 publications

References 19 publications

Group-Based Recurrent Neural Networks for POI Recommendation

Group-Based Recurrent Neural Networks for POI Recommendation

A roadmap for semi-automatically extracting predictive and clinically meaningful temporal features from medical data for predictive modeling

Predictive Multi-level Patient Representations from Electronic Health Records

Contact Info

Product

Resources

About