Catalogue Search | MBRL
Are you sure you want to remove the book from the shelf?
{{itemTitle}}
621
result(s) for
"Lim, Ee"
Sort by:
Computational trust models and machine learning
\"This book provides an introduction to computational trust models from a machine learning perspective. After reviewing traditional computational trust models, it discusses a new trend of applying formerly unused machine learning methodologies, such as supervised learning. The application of various learning algorithms, such as linear regression, matrix decomposition, and decision trees, illustrates how to translate the trust modeling problem into a (supervised) learning problem. The book also shows how novel machine learning techniques can improve the accuracy of trust assessment compared to traditional approaches\"-- Provided by publisher.
Book
Non-binary evaluation of next-basket food recommendation
Next-basket recommendation (NBR) is a recommendation task that predicts a basket or a set of items a user is likely to adopt next based on his/her history of basket adoption sequences. It enables a wide range of novel applications and services from predicting next basket of items for grocery shopping to recommending food items a user is likely to consume together in the next meal. Even though much progress has been made in the algorithmic NBR research over the years, little research has been done to broaden knowledge about the evaluation of NBR methods, which is largely based on the offline evaluation experiments and binary relevance paradigm. Specifically, we argue that recommended baskets which are more similar to ground truth baskets are better recommendations than those that share little resemblance to the ground truth, and therefore, they should be granted some partial credits. Based on this notion of non-binary relevance assessment, we propose new evaluation metrics for NBR by adapting and extending similarity metrics from natural language processing (NLP) and text classification research. To validate the proposed metrics, we conducted two user studies on the next-meal food recommendation using numerous state-of-the-art NBR methods in both online and offline evaluation settings. Our findings show that the offline performance assessment based on the proposed non-binary evaluation metrics is more representative of the online evaluation performance than that of the standard evaluation metrics.
Journal Article
Network data mining and analysis
\"Consider an online social networking site with millions of members in which members have the opportunity to befriend one another, send messages to each other, and post content on the site. Facebook, LinkedIn, and Twitter are examples of such sites. To make sense of data from these sites, we resort to social media mining to answer the following questions: 1. What are social communities in bipartite graphs and signed graphs? 2. How robust are the networks? How can we apply the robustness of networks? 3. How can we find identical social users across heterogeneous social networks? Social media shatters the boundaries between the real world and the virtual world. We can now integrate social theories with computational methods to study how individuals interact with each other and how social communities form in bipartite and signed networks. The uniqueness of social media data calls for novel data mining techniques that can effectively handle user generated content with rich social relations. The study and development of these new techniques are under the purview of social media mining, an emerging discipline under the umbrella of data mining. Social Media Mining is the process of representing, analyzing, and extracting actionable patterns from social media data\"-- Provided by publisher.
Book
A fully humanized IgG-like bispecific antibody for effective dual targeting of CXCR3 and CCR6
Chemokines and their receptors are pivotal for the trafficking of leukocytes during immune responses, and host defense. However, immune cell migration also contributes to a wide variety of autoimmune and chronic inflammatory diseases. Compelling evidence suggests that both CXCR3 and CCR6 chemokine receptors play crucial roles in the migration of pathological Th1 and Th17 cells during the course of certain inflammatory diseases. The use of two or more receptors by pathogenic cells may explain why targeting of individual receptors has proven disappointing in the clinic. We therefore hypothesized that simultaneous targeting of both CXCR3 and CCR6 with a bispecific antibody (BsAb) might result in decreased chemotaxis and/or specific depletion of pro-inflammatory T cell subsets. In this study, we designed and characterized a fully humanized BsAb. We show that the BsAb binds to both chemokine receptors, as demonstrated by Flow Cytometry and Surface Plasmon Resonance analysis. Furthermore, we demonstrate that the BsAb effectively blocks cell chemotaxis and induces specific antibody-dependent cell-mediated cytotoxicity (ADCC) in vitro. Therefore, we propose that dual targeting of CXCR3 and CCR6 with a fully humanized BsAb may display a potent interventional approach for the treatment of inflammatory and autoimmune diseases.
Journal Article
Functional and structural analysis of non-synonymous single nucleotide polymorphisms (nsSNPs) in the MYB oncoproteins associated with human cancer
MYB proteins are highly conserved DNA-binding domains (DBD) and mutations in MYB oncoproteins have been reported to cause aberrant and augmented cancer progression. Identification of MYB molecular biomarkers predictive of cancer progression can be used for improving cancer management. To address this, a biomarker discovery pipeline was employed in investigating deleterious non-synonymous single nucleotide polymorphisms (nsSNPs) in predicting damaging and potential alterations on the properties of proteins. The nsSNP of the MYB family;
MYB
,
MYBL1
, and
MYBL2
was extracted from the NCBI database. Five in silico tools (PROVEAN, SIFT, PolyPhen-2, SNPs&GO and PhD-SNP) were utilized to investigate the outcomes of nsSNPs. A total of 45 nsSNPs were predicted as high-risk and damaging, and were subjected to PMut and I-Mutant 2.0 for protein stability analysis. This resulted in 32 nsSNPs with decreased stability with a DDG score lower than − 0.5, indicating damaging effect. G111S, N183S, G122S, and S178C located within the helix-turn-helix (HTH) domain were predicted to be conserved, further posttranslational modifications and 3-D protein analysis indicated these nsSNPs to shift DNA-binding specificity of the protein thus altering the protein function. Findings from this study would help in the field of pharmacogenomic and cancer therapy towards better intervention and management of cancer.
Journal Article
Hungry bone syndrome after parathyroidectomy in end-stage renal disease patients: review of an alkaline phosphatase-based treatment protocol
AimHyperparathyroidism in chronic kidney disease–mineral and bone disorder is associated with significant morbidity and mortality. Parathyroidectomy is widely carried out as treatment despite complications such as hypocalcaemia post-surgery. Our centre has been using an ALP-based protocol to replace calcium postoperatively to prevent hypocalcaemia. We aim to describe and audit our calcium replacement protocol post-parathyroidectomyMethodsWe, retrospectively, analyse 167 end-stage kidney disease patients who had parathyroidectomy with auto-implantation in Singapore General Hospital between January 2008 and December 2013. Their calcium replacement postoperatively was initiated upon patient arrival back in ward on the same day of surgery based on their pre-op ALP prior to occurrence of hypocalcaemia. Patient demographics, surgical and laboratory parameters were reviewed from medical records. Changes in calcium postoperatively were reported to look for incidence of calcium derangement.ResultsMean calcium levels between pre-operation day and post-operation day 7 ranged from 2.31 to 2.70 mmol/L. Decline in serum calcium was common in all patients prior to starting calcium replacement. Eighteen patients (10.9%) experienced hypocalcaemia immediately post-operation prior to commencement of IV calcium replacement. Patients with immediate post-operation hypocalcaemia had lower pre-operation calcium but higher pre-operation alkaline phosphatase (ALP) and pre-operation intact parathyroid hormone. Hypercalcaemia is common likely from aggressive IV calcium replacement using the protocol. The average length of stay for patients prior to calcium stabilization and discharge was 9 days.ConclusionImplementation of an ALP-based prophylactic calcium replacement protocol with daily serum calcium monitoring can ameliorate severe hypocalcaemia post-parathyroidectomy
Journal Article
A transformer framework for generating context-aware knowledge graph paths
Contextual Path Generation (CPG) refers to the task of generating knowledge path(s) between a pair of entities mentioned in an input textual context to determine the semantic connection between them. Such knowledge paths, also called contextual paths, can be very useful in many advanced information retrieval applications. Nevertheless, CPG involves several technical challenges, namely, sparse and noisy input context, missing relations in knowledge graphs, and generation of ill-formed and irrelevant knowledge paths. In this paper, we propose a transformer-based model architecture. In this approach, we leverage a mixture of pre-trained word and knowledge graph embeddings to encode the semantics of input context, a transformer decoder to perform path generation controlled by encoded input context and head entity to stay relevant to the context, and scaling methods to sample a well-formed path. We evaluate our proposed CPG models derived using the above architecture on two real datasets, both consisting of Wikinews articles as input context documents and ground truth contextual paths, as well as a large synthetic dataset to conduct larger-scale experiments. Our experiments show that our proposed models outperform the baseline models, and the scaling methods contribute to better quality contextual paths. We further analyze how CPG accuracy can be affected by different amount of context data, and missing relations in the knowledge graph. Finally, we demonstrate that an answer model for knowledge graph questions adapted for CPG could not perform well due to the lack of an effective path generation module.
Journal Article
Elevated Circulating Osteoprotegerin and Renal Dysfunction Predict 15-Year Cardiovascular and All-Cause Mortality: A Prospective Study of Elderly Women
Data on the predictive role of estimated glomerular filtration rate (eGFR) and osteoprotegerin (OPG) for cardiovascular (CVD) and all-cause mortality risk have been presented by our group and others. We now present data on the interactions between OPG with stage I to III chronic kidney disease (CKD) for all-cause and CVD mortality.
The setting was a 15-year study of 1,292 women over 70 years of age initially randomized to a 5-year controlled trial of 1.2 g of calcium daily. Serum OPG and creatinine levels with complete mortality records obtained from the Western Australian Data Linkage System were available. Interactions were detected between OPG levels and eGFR for both CVD and all-cause mortality (P < 0.05). Compared to participants with eGFR ≥60 ml/min/1.73 m2 and low OPG, participants with eGFR of <60 ml/min/1.73 m2 and elevated OPG had a 61% and 75% increased risk of all-cause and CVD mortality respectively (multivariate-adjusted HR, 1.61; 95% CI, 1.27-2.05; P < 0.001 and HR, 1.75; 95% CI, 1.22-2.55; P = 0.003). This relationship with mortality was independent of decline in renal function (P<0.05). Specific causes of death in individuals with elevated OPG and stage III CKD highlighted an excess of coronary heart disease, renal failure and chronic obstructive pulmonary disease deaths (P < 0.05).
The association between elevated OPG levels with CVD and all-cause mortality was more evident in elderly women with poorer renal function. Assessment of OPG in the context of renal function may be important in studies investigating its relationship with all-cause and CVD mortality.
Journal Article
Promoting Sustainable Urban Mobility: Factors Influencing E-Bike Adoption in Henan Province, China
This study examines the key factors influencing e-bike adoption and explores how advancing e-bike usage in Henan Province, China, can foster sustainable urban transportation and contribute to urban environmental preservation. Utilizing data from an online survey, binary logistic regression analyzes the impact of socio-demographic characteristics, perceived advantages, neighborhood environmental attributes, and vehicle ownership on e-bike usage. The findings indicate that socio-demographic factors, such as family size and occupation, significantly influence adoption, with workmen more likely than office workers to choose e-bikes. Cost savings emerged as the primary motivator for e-bike use, overshadowing environmental concerns, which unexpectedly negatively affected usage patterns. However, the presence of supportive infrastructure—particularly charging stations and dedicated lanes—proves crucial for promoting e-bike usage, highlighting the importance of accessible, environmentally supportive urban design. Vehicle ownership characteristics further illuminate how access to e-bikes correlates with regular usage. These findings suggest that, beyond cost efficiency, targeted awareness campaigns and strategic infrastructure improvements are essential for embedding e-bikes into sustainable urban transport systems. By fostering adoption and supporting e-bike infrastructure, cities can significantly reduce urban pollution, improve air quality, and advance toward sustainable mobility goals in Henan Province and beyond.
Journal Article
On measuring network robustness for weighted networks
Network robustness measures how well network structure is strong and healthy when it is under attack, such as vertices joining and leaving. It has been widely used in many applications, such as information diffusion, disease transmission, and network security. However, existing metrics, including node connectivity, edge connectivity, and graph expansion, can be suboptimal for measuring network robustness since they are inefficient to be computed and cannot directly apply to the weighted networks or disconnected networks. In this paper, we define the R-energy as a new robustness measurement for weighted networks based on the method of spectral analysis. R-energy can cope with disconnected networks and is efficient to compute with a time complexity of O(|V|+|E|), where V and E are sets of vertices and edges in the network, respectively. Our experiments illustrate the rationality and efficiency of computing R-energy: (1) Removal of high degree vertices reduces network robustness more than that of random or small degree vertices; (2) it takes as little as 120 s to compute for a network with about 6M vertices and 33M edges. We can further detect events occurring in a dynamic Twitter network with about 130K users and discover interesting weekly tweeting trends by tracking changes to R-energy.
Journal Article