Directed Acyclic Graph Research Articles

Causal structure learning for high-dimensional non-stationary time series

Learning the causal structure of high-dimensional non-stationary time series can help in understanding the data generation mechanism, which is a crucial task in machine learning. However, current causal discovery methods for high-dimensional non-stationary time series face several challenges, including the inability to effectively capture non-stationarity, failure to ensure acyclicity of causal graphs, and reliance on subjective threshold definitions, leading to suboptimal algorithm performance. To address these challenges, we introduce a novel Causal Structure Learning model for High-dimensional Non-stationary Time Series (CSL-HNTS). Firstly, this model presents a graph neural network to model the non-stationarity of time series. Secondly, it introduces a novel Directed Acyclic Graph (DAG) sampling method that transforms the space of DAGs into a continuous space, enabling the search for causal graphs within this continuous space to ensure acyclicity. Finally, the model proposes an automatic threshold definition method, without prior knowledge, to convert the weighted adjacency matrix into the Boolean adjacency matrix of the causal graph, thereby avoiding time-consuming postprocessing steps. The proposed approach is validated using simulation datasets and two real datasets, and is benchmarked against current state-of-the-art methods and ablation experiments. The results demonstrate a significant improvement over existing methods, highlighting the efficacy of the proposed model.

Network-based representations and dynamic discrete choice models for multiple discrete choice analysis

In many choice modeling applications, consumer demand is frequently characterized as multiple discrete, which means that consumer choose multiple items simultaneously. The analysis and prediction of consumer behavior in multiple discrete choice situations pose several challenges. In this paper, to address this, we propose a random utility maximization (RUM) based model that considers each subset of choice alternatives as a composite alternative, where individuals choose a subset according to the RUM framework. While this approach offers a natural and intuitive modeling approach for multiple-choice analysis, the large number of subsets of choices in the formulation makes its estimation and application intractable. To overcome this challenge, we introduce directed acyclic graph (DAG) based representations of choices where each node of the DAG is associated with an elemental alternative and additional information such as the number of selected elemental alternatives. Our innovation is to show that the multi-choice model is equivalent to a recursive route choice model on the DAG, leading to the development of new efficient estimation algorithms based on dynamic programming. In addition, the DAG representations enable us to bring some advanced route choice models to capture the correlation between subset choice alternatives. Numerical experiments based on synthetic and real datasets show many advantages of our modeling approach and the proposed estimation algorithms.

A Causal Framework for the Comparability of Latent Variables

Measurement invariance (MI) describes the equivalence of measurement models of a construct across groups or time. When comparing latent means, MI is often stated as a prerequisite of meaningful group comparisons. The most common way to investigate MI is multi-group confirmatory factor analysis (MG-CFA). Although numerous guides exist, a recent review showed that MI is rarely investigated in practice. We argue that one reason might be that the results of MG-CFA are uninformative as to why MI does not hold between groups. Consequently, under this framework, it is difficult to regard the study of MI as an interesting and constructive step in the modeling process. We show how directed acyclic graphs (DAGs) from the causal inference literature can guide researchers in reasoning about the causes of non-invariance. For this, we first show how DAGs for measurement models can be translated into path diagrams used in the linear structural equation model (SEM) literature. We then demonstrate how insights gained from this causal perspective can be used to explicitly model encoded causal assumptions with moderated SEMs, allowing for a more enlightening investigation of MI. Ultimately, our goal is to provide a framework in which the investigation of MI is not deemed a “gateway test” that simply licenses further analyses. By enabling researchers to consider MI as an interesting part of the modeling process, we hope to increase the prevalence of investigations of MI altogether.

Enhancing smart city operation management: Integrating energy systems with a subway synergism hub

This paper is centered on establishing a secure framework for the optimal concurrent operation of a smart city, encompassing transportation, water, heat, electrical, and cooling energy systems. The studied smart city includes the microgrid, smart transportation system (STS), energy hub (EH) and smart grid. In this regard, a subway synergism hub (SSH) as a new non-energy system is added to the smart city with the aim of serving the subway's water, heat, electrical and cooling demands as well as diminishing the operation cost of the smart city. The EH within the SSH cooperated with a desalination unit is considered to supply the subway's stations water demand by using the sea water. The investigation of the optimal allocation of the SSH unit for reducing the cost of smart city operation is also conducted by introducing a novel intelligent priority selection (IPS) analytical algorithm. In comparison to common meta-heuristic algorithms for allocation problems, the accurate optimal solution can be found in low runtime by the IPS algorithm. To achieve an accurate model of the smart city, directed acyclic graph (DAG) based blockchain approach is provided which can enhance the data and energy exchanges security within the smart city. This research paper introduces a security framework deployed in a smart city setting to establish a secure platform for energy transactions. The findings validate the effectiveness of this model and highlight the value of the IPS method. The effectiveness of the suggested approach has been assessed using the smart city system is comprised of various sections, including EVs, smart grid, microgrid, and SSH, demonstrating the credibility and accuracy of this study.

Higher consumption of ultra-processed foods is associated with obesity and abdominal obesity in socially vulnerable Brazilian women.

This study aimed to assess the consumption of ultra-processed foods (UPF) and identify their association with obesity and abdominal obesity in adult women of reproductive age living insituations of social vulnerability in Maceió, Northeastern Brazil. This was a cross-sectional study carried out between October 2020 and May 2021. An anthropometric evaluation was carried out to assess obesity and abdominal obesity. A dietary assessment was also conducted using a 24-h food recall to determine the calorie intake from UPF. To estimate intra-individual variability in food consumption, the probabilistic Multiple Source Method was used. These data in the form of tertiles were used to analyse the association between the consumption of UPF and obesity and abdominal obesity. Logistic regressions were used to analyse the association. A directed acyclic graph (DAG) was created for this analysis. This study included 1702 women of which 53.7% were 31 years old or older, and 74.2% lived in poverty. It identified that 36.5% and 38.1% of the women had obesity and abdominal obesity, respectively, and that an average of 33.8% of calories consumed came from UPF. In the analysis of association guided by the DAG, it was observed that women with a high-calorie intake from UPF had a 1.3 times higher probability of being obese. It was also observed that women with a moderate and high-calorie intake from UPF were 1.4 and 1.3 times more likely, respectively, to have abdominal obesity. Thus, it can be concluded that socially vulnerable women in Brazil have a relatively high consumption of UPF and that this condition increases the probability of obesity in this population group.

A copula-based secured intelligent dynamic-static energy community transportation system for smart cities

This paper addresses a secured co-dynamic model for the energy management of Electrical Vehicles (EVs) within the real community transportation system (RCTS). The proposed model aims to facilitate interoperability among mobile energy resources within the smart city, enabling the RCTS to model the co-dynamic-static transportation systems (TSs) simultaneously. The energy management model within the traffic flow system focuses on dynamic assignment, considering the power consumption associated with the density of moving vehicles. EVs play a key role in economically managing energy in both static and dynamic behaviors within charging stations while aligning with the current traffic flow.To enhance data security within the smart city ecosystem, a directed acyclic graph (DAG)-based decentralized cyber security approach is recommended. This approach ensures that data transactions involving mobile energy resources are secured against cyber-attacks through the use of public, private, and transaction blocks. Additionally, an uncertainty-based copula function is presented to create a precise management environment within the smart city. The results indicate that the proposed model for transportation energy resources tends to reduce energy costs by optimally controlling energy consumption within traffic flow, compared to normal conditions.

Impact of therapeutic plasmapheresis on the duration of organ failure in patients with hypertriglyceridemia-associated acute pancreatitis

BackgroundPlasmapheresis is widely used for severe hypertriglyceridemia-associated acute pancreatitis (HTG-AP) to remove excessive triglycerides from plasma. This study aimed to evaluate whether plasmapheresis could improve the duration of organ failure in HTG-AP patients.MethodsWe analyzed a cohort of patients from a multicenter, prospective, long-running registry (the PERFORM) collecting HTG-AP patients admitted to the study sites within 72 h from the onset of symptoms. This study was based on data collected from November 2020 to March 2023. Patients who had organ failure at enrollment were involved in the analyses. The primary outcome was time to organ failure resolution within 14 days. Multivariable Cox regression model was used to evaluate the association between plasmapheresis and time to organ failure resolution. Directed acyclic graph (DAG) was used to identify potential confounders.ResultsA total of 122 HTG-AP patients were included (median [IQR] sequential organ failure assessment (SOFA) score at enrollment, 3.00 [2.00–4.00]). Among the study patients, 46 underwent plasmapheresis, and 76 received medical treatment. The DAG revealed that baseline serum triglyceride, APACHE II score, respiratory failure, cardiovascular failure, and renal failure were potential confounders. After adjusting for the selected confounders, there was no significant difference in time to organ failure resolution between patients undergoing plasmapheresis and those receiving exclusive medical treatment (HR = 1.07; 95%CI 0.68–1.68; P = 0.777). Moreover, the use of plasmapheresis was associated with higher ICU requirements (97.8% [45/46] vs. 65.8% [50/76]; OR, 19.33; 95%CI 2.20 to 169.81; P = 0.008).ConclusionsIn HTG-AP patients with early organ failure, plasmapheresis was not associated with accelerated organ failure resolution compared to medical treatment but may be associated with more ICU admissions.Trial registration: The PERFORM study was registered in the Chinese Clinical Trial Registry (ChiCTR2000039541). Registered 30 October 2020.

Mapping network connection and direction among symptoms of depression and anxiety in patients with chronic gastritis.

Regarding neurophysiological and developmental findings, anxiety and depression are usual comorbidities of gastritis patients. However, research related to anxiety and depression among chronic gastritis patients was conducted on the disease level while ignoring symptoms. Hence, we rendered the network approach to reveal the symptoms of anxiety and depression among chronic gastritis patients. Three hundred and sixty-nine chronic gastritis patients (female = 139, Mage = 55.87 years) were asked to complete the Self-Rating Anxiety Scale and Self-Rating Depression Scale. Three symptom networks and one directed acyclic graph (DAG) network were formed. First, in the anxiety network of chronic gastritis patients, dizziness was the most influential symptom. In the depression network of chronic gastritis patients, depressed affect and psychomotor retardation were the influential symptoms. Second, panic, easy fatiguability, weakness, palpitation, depressed affect, tachycardia, fatigue, and psychomotor agitation bridged the anxiety-depression network of chronic gastritis patients. Third, DAG networks showed that anxiousness and hopelessness could trigger other symptoms in the anxiety-depression networks of chronic gastritis patients. The current study provided insightful information on patients with chronic gastritis by examining the structures of symptoms.

Distributed matrix computing system for big data

In order to solve the problem of low computing efficiency in big data analysis and model construction, this paper intended to deeply explore the big data analysis programming model, DAG (Directed Acyclic Graph) and other contents, and on this basis, it adopted a distributed matrix computing system Octopus for big data analysis. Octopus is a universal matrix programming framework that provides a programming model based on matrix operations, which can conveniently analyze and process large-scale data. By using Octopus, users can extract functions and data from multiple platforms and operate through a unified matrix operation interface. The distributed matrix representation and storage layer can design data storage formats for distributed file systems. Each computing platform in OctMatrix provides its own matrix library, and it provides a matrix library written in R language for the above users. SymboMatrix provides a matrix interface to OctMatrix that is consistent with OctMatrix. However, SymboMatrix also retains the flow diagram for matrix operations in the process, and it also supports logical and physical optimization of the flow diagram on a DAG. For the DAG computational flow graph generated by SymbolMatrix, this paper divided it into two parts: logical optimization and physical optimization. This paper adopted a distributed file system based on line matrix, and obtained the corresponding platform matrix by reading the documents based on line matrix. In the evaluation of system performance, it was found that the distributed matrix computing system had a high computing efficiency, and the average CPU (central processing unit) usage reached 70%. This system can make full use of computing resources and realize efficient parallel computing.

Efficient Optimization approach in WSN using Multicast Forwarding Scheme in 6LowPAN Environment

The rapidly growing network of devices that are connected known as the Internet of Things, also referred to as IoT, is designed to collect and exchange data. These devices are integrated with sensors, software, and other technologies. An IPv6-based routing system called 6LowPAN was created especially for Internet of Things devices with little resources. However, because multicast packets must be replicated to every multicast group member, the conventional multicast forwarding technique in 6LowPAN may be wasteful. This can result in excessive energy usage and network traffic, particularly in large-scale Internet of Things implementations. An optimization process is presented here using a multicast forwarding scheme. In the proposed forwarding scheme, nodes can send multicast packets up and down at Low Power and Lossy networks over the 6LowPAN environment. Should the sink node be on the root node, it should immediately request that other nodes join the multicast messaging network of things. If otherwise, an ICMPv6 packet will be sent towards the root node in the Destination-oriented Directed-Acyclic Graphs (DODAG) tree, which expands the multicast messages in the first source's interest. In multi-bounce forward straight-line topology, when the DODAG network root is the source of multicast congestion, it exhibits a similar ratio of packet delivery and delay from end to end with a minimal memory overhead. When traffic via multicast originates from the most important rank node in a backward straight-line topology, this paper focuses on the efficiency of the multicast network at 6LowPAN.

Causal diagrams for civil and structural engineers

Causal diagrams are logic and graphical tools that depict assumptions about presumed causal relations. Such diagrams have proven effective in tackling a variety of problems in social sciences and epidemiology research yet remain foreign to civil engineers. Unlike the traditional means of examining relationships via multivariable regression, causal diagrams can identify the presence of confounders, colliders, and mediators. Thus, this paper hopes to introduce the big ideas behind causal diagrams (specifically, directed acyclic graphs (DAGs)) and how to create and apply such diagrams to several civil engineering problems. Findings from the presented case studies indicate that civil engineers can successfully use causal diagrams to improve their understanding of complex causation relations, thereby accelerating research and practical efforts.

Towards secure and efficient integration of blockchain and 6G networks.

The future of communication systems is undergoing a transformative shift towards intelligence, efficiency, and flexibility. Presently, the amalgamation of blockchain technology and the sixth-generation mobile communication network (6G) has garnered significant attention, as their fusion is poised to profoundly impact the digital economy and society at large. However, the convergence of blockchain and 6G networks poses challenges pertaining to security and performance. In this article, we propose an approach based on the design of secure mechanisms and performance optimization to delve into the key issues surrounding the integration of blockchain and 6G networks from both security and performance perspectives. Specifically, we first introduce the application scenarios of 6G networks and blockchain's empowerment of them to highlight the necessity of combining blockchain technology with 6G. Subsequently, in order to ensure the security of communication and data transmission between blockchain and 6G networks, we have investigated the design requirements for security mechanisms. Furthermore, we discuss the efficient realization of the amalgamation between blockchain and 6G networks by proposing a solution based on Directed Acyclic Graph (DAG) for blockchain's asynchronous consensus protocol, alongside optimization strategies for storage and communication to meet the desired characteristics and requirements of 6G networks. Lastly, we provide valuable research directions that serve as references and guidance for the future development of the integration between blockchain and 6G networks.

Incorporating equity, diversity, and inclusion into the epidemiology and biostatistics curriculum: A workshop report and implementation strategies recommendations.

There is an obligation among those teaching epidemiology to incorporate principles of equity, diversity, and inclusion (EDI) into the curriculum. While there is a well-established literature related to teaching epidemiology, this literature rarely addresses critical aspects of EDI. To our knowledge, there is no working group or central point of discussion and learning for incorporating EDI into epidemiology teaching in Canada. To address this gap, we convened a workshop entitled "Incorporating EDI into the epidemiology and biostatistics curriculum and classroom." The workshop discussed nine strategies to incorporate EDI in the epidemiology curriculum: positionality (or reflexivity) statements; opportunities for feedback; land acknowledgements; clarifying the purpose of collecting data on race and ethnicity, sex and gender, Indigeneity; acknowledgingthat race/ethnicity is asocial construct, not a biological variable; describing incidence and prevalence of disease; demonstrating explicit bias using directed acyclic graphs (DAGs); critical appraisal of study population diversity; and admission criteria and considerations. Key take-aways from the workshop were the need to be more intentional when determining the validity of evidence, particularly with respect to historical context and the need to recognize that there is no single solution that will address EDI.

Environmental factors associated with Escherichia coli concentration at freshwater beaches on Lake Winnipeg, Manitoba, Canada

At many public beaches, routine monitoring of beach water quality using fecal indicator bacteria is conducted to evaluate the risk of recreational water illness. Results from water sample analysis can take over 24-hr, which may no longer accurately reflect current water quality conditions. This study aimed to assess which combination of environmental factors best predicts fecal contamination (E. coli) levels at two of the most popular beaches on Lake Winnipeg, Manitoba (Gimli and Grand Beach), by linking water quality data and publicly available environmental data from 2007 to 2021. We developed separate mixed effects models for each beach for two outcomes, linear (continuous log-transformed E. coli concentration) and categorical (200 CFU/100 ml threshold), to explore differences in the predictors of E. coli concentrations and exceedances of the provincial health risk threshold, respectively. We used a Directed Acyclic Graph to choose which predictor variables to include in the models. For both beaches, we identified clustering of the E. coli outcomes by year, suggesting year-specific variation. We also determined that extreme weather days, with higher levels of rainfall in the preceding 48-hr, previous day average air temperature, and previous day E. coli concentration could result in a higher probability of E. coli threshold exceedances or higher concentrations in the water bodies. In Grand Beach, we identified that days with lower average UV levels in the previous 24-hr and antecedent dry days could result in a higher probability of E. coli threshold exceedances or higher concentrations. The findings can inform possible trends in other freshwater settings and be used to help develop real-time recreational water quality predictive models to allow more accurate beach management decisions and warrant enhancement of beach monitoring programs for extreme weather events as part of the climate change preparedness efforts.

Invited commentary: where do the causal DAGS come from?

How do we construct our causal directed acyclic graphs (DAGs)-for example, for life-course modeling and analysis? In this commentary, I review how the data-driven construction of causal DAGs (causal discovery) has evolved, what promises it holds, and what limitations or caveats must be considered. I find that expert- or theory-driven model-building might benefit from some more checking against the data and that causal discovery could bring new ideas to old theories.

Exploring a Potential Interaction Between the Effect of Specific Maternal Smoking Patterns and Comorbid Antenatal Depression in Causing Postpartum Depression.

To explore a potential interaction between the effect of specific maternal smoking patterns and the presence of antenatal depression, as independent exposures, in causing postpartum depression (PPD). This case-control study of participants with singleton term births (N = 51220) was based on data from the 2017-2018 Pregnancy Risk Assessment Monitoring System. Multivariable log-binomial regression models examined the main effects of smoking patterns and self-reported symptoms of antenatal depression on the risk of PPD on the adjusted risk ratio (aRR) scale and tested a two-way interaction adjusting for covariates selected in a directed acyclic graph (DAG). The interaction effects were measured on the additive scale using relative excess risk due to interaction (RERI), the attributable proportion of interaction (AP), and the synergy index (SI). Causal effects were defined in a counterfactual framework. The E-value quantified the potential impact of unobserved/unknown covariates, conditional on observed covariates. Among 6841 women in the sample who self-reported PPD, 35.7% also reported symptoms of antenatal depression. Out of 3921 (7.7%) women who reported smoking during pregnancy, 32.6% smoked at high intensity (≥10 cigarettes/day) in all three trimesters and 36.6% had symptoms of antenatal depression. The main effect of PPD was the strongest for women who smoked at high intensity throughout pregnancy (aRR 1.65; 95% CI: 1.63, 1.68). A synergistic interaction was detected, and the effect of all maternal smoking patterns was augmented, particularly in late pregnancy for Increasers and Reducers. Strong associations and interaction effects between maternal smoking patterns and co-occurring antenatal depression support smoking prevention and cessation interventions during pregnancy to lower the likelihood of PPD.

On Low-Rank Directed Acyclic Graphs and Causal Structure Learning.

Despite several advances in recent years, learning causal structures represented by directed acyclic graphs (DAGs) remains a challenging task in high-dimensional settings when the graphs to be learned are not sparse. In this article, we propose to exploit a low-rank assumption regarding the (weighted) adjacency matrix of a DAG causal model to help address this problem. We utilize existing low-rank techniques to adapt causal structure learning methods to take advantage of this assumption and establish several useful results relating interpretable graphical conditions to the low-rank assumption. Specifically, we show that the maximum rank is highly related to hubs, suggesting that scale-free (SF) networks, which are frequently encountered in practice, tend to be low rank. Our experiments demonstrate the utility of the low-rank adaptations for a variety of data models, especially with relatively large and dense graphs. Moreover, with a validation procedure, the adaptations maintain a superior or comparable performance even when graphs are not restricted to be low rank.

PICO: Pipeline Inference Framework for Versatile CNNs on Diverse Mobile Devices

Distributing the inference of convolutional neural network (CNN) to multiple mobile devices has been studied in recent years to achieve real-time inference without losing accuracy. However, how to map CNN to devices remains a challenge. On the one hand, scheduling the workload of state-of-the-art CNNs with multiple devices is NP-Hard because the structures of CNNs are directed acyclic graphs (DAG) rather than simple chains. On the other hand, distributing the inference workload suffers from expensive communication and unbalanced computation due to the wireless environment and heterogeneous devices. This paper presents PICO, a pipeline cooperation framework to accelerate the inference of versatile CNNs on diverse mobile devices. At its core, PICO features: (1) a generic graph partition algorithm that considers the characteristics of any given CNN and orchestrates it into a list of model pieces with suitable granularity, and (2) a many-to-many mapping algorithm that produces the best pipeline configuration for heterogeneous devices. In our experiment with <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$2 \sim 8$</tex-math></inline-formula> Raspberry-Pi devices, the throughput can be improved by <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$1.8 \sim 6.8 \times$</tex-math></inline-formula> under different CPU frequencies.

On State Transition Probability and Performance of Direct Acyclic Graph Based Ledgers

Direct acyclic graph (DAG) based ledgers with multi-chain structures aim to solve the technical bottlenecks associated with classical blockchain technologies in the Internet of Things (IoT). The basic working principle of DAG-based ledgers is to validate new transactions by previous transactions in order to be added to the system. During the tip selection process in the unsteady regime, the state transition probability refers to the probability of a transaction changing from the initial state to an arbitrary state. The state transition probability plays an indispensable role in the performance and security analysis of the IoT relying on DAG-based ledgers. In this paper, we derive the exact expression and an approximate expression of the state transition probability, which both are in closed form. In addition, we propose and analyze three performance metrics, i.e., the expected cumulative weight, the expected number of steps, and the confirmation failure probability, which are derived from the state transition probability and greatly enrich the performance analysis and evaluation of the IoT. Markov chain Monte Carlo (MCMC) simulations are carried out to verify the derived analytical results and provide insight into the IoT using DAG-based ledgers.

Network Analysis of the Relationships Between Symptoms, Functioning, and Sleep/Fatigue in Colorectal Cancer Patients

IntroductionAmong the various patient experiences, cancer-related fatigue and sleep disturbances emerge as pivotal aspects that can substantially impact individuals’ quality of life. There exists a relative scarcity of research focusing on the intricate relationship between symptoms, functioning, fatigue, and sleep disturbances in colorectal cancer (CRC) patients.ObjectivesIn this context, the current research endeavors to apply advanced statistical methodologies to elucidate the complex relationships between symptoms, functioning, fatigue, and sleep disturbances. By exploring the intricate web of patient characteristics, clinical factors, psychosocial elements, this study aims to construct a holistic model that not only captures the nuances of colorectal cancer patients’ experiences but also uncovers potential avenues for intervention and support.MethodsIn our cross sectional study, we administered the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire Core 30 (EORTC QLQ-C30), the Quality of Life Questionnaire Colorectal Cancer Module (QLQ-CR29) to 987patients who were surgically-treated for CRC from the tertiary hospital from 2013 through 2018. To confirm the relationship between symptoms of CRC patients, univariable logistic regression was used to examine the potential relationship between independent variables and the occurrence of fatigue and sleep disorders. Least Absolute Shrinkage and Selection Operator (Lasso) was used for variable selection. The selected variables were then applied to a multivariate logistic regression analysis to examine the most influential predictors of fatigue and sleep disturbance. Finally, gaussian graphical models (GGM) were used to identify potential interactions between characteristics, symptoms, functioning, with fatigue, and sleep disturbances in CRC. In this study, Directed Acyclic Graph (DAG) was used to identify causal dependancy and path of variables.Results About 10.4% of study participants reported experiencing fatigue. Sleep problems were reported by 15.8% of the study participants. Multivariable logistic regression analysis using Lasso showed that sleep problem (odds ratio [OR]=2.34; 95% CI, 1.03-5.31), physical, role, and emotional functioning, pain, dyspnoea, and appetite loss were significant predictors of fatigue, while emotional functioning, dyspnoea, and appetite loss were significant predictors of sleep problem. The variables that were directly linked to fatigue were role functioning, emotional functioning, dyspnoea, appetite loss, body image and trouble with taste. The variables that were directly linked to sleep problem were emotional functioning and appetite loss.ConclusionsIn conclusion, there were complex relationships between symptoms, functioning, fatigue, and sleep disturbances. The symptom network of CRC patients showed different patterns toward fatigue and sleep.Disclosure of InterestNone Declared