RESUMO
Next-basket recommendation considers the problem of recommending a set of items into the next basket that users will purchase as a whole. In this paper, we develop a novel mixed model with preferences, popularities and transitions (M2) for the next-basket recommendation. This method models three important factors in next-basket generation process: 1) users' general preferences, 2) items' global popularities and 3) transition patterns among items. Unlike existing recurrent neural network-based approaches, M2 does not use the complicated networks to model the transitions among items, or generate embeddings for users. Instead, it has a simple encoder-decoder based approach (ed-Trans) to better model the transition patterns among items. We compared M2 with different combinations of the factors with 5 state-of-the-art next-basket recommendation methods on 4 public benchmark datasets in recommending the first, second and third next basket. Our experimental results demonstrate that M2 significantly outperforms the state-of-the-art methods on all the datasets in all the tasks, with an improvement of up to 22.1%. In addition, our ablation study demonstrates that the ed-Trans is more effective than recurrent neural networks in terms of the recommendation performance. We also have a thorough discussion on various experimental protocols and evaluation metrics for next-basket recommendation evaluation.
RESUMO
Sequential recommendation aims to identify and recommend the next few items for a user that the user is most likely to purchase/review, given the user's purchase/rating trajectories. It becomes an effective tool to help users select favorite items from a variety of options. In this manuscript, we developed hybrid associations models (HAM) to generate sequential recommendations. using three factors: 1) users' long-term preferences, 2) sequential, high-order and low-order association patterns in the users' most recent purchases/ratings, and 3) synergies among those items. HAM uses simplistic pooling to represent a set of items in the associations, and element-wise product to represent item synergies of arbitrary orders. We compared HAM models with the most recent, state-of-the-art methods on six public benchmark datasets in three different experimental settings. Our experimental results demonstrate that HAM models significantly outperform the state of the art in all the experimental settings. with an improvement as much as 46.6%. In addition, our run-time performance comparison in testing demonstrates that HAM models are much more efficient than the state-of-the-art methods. and are able to achieve significant speedup as much as 139.7 folds.
RESUMO
MOTIVATION: Graph embedding learning that aims to automatically learn low-dimensional node representations, has drawn increasing attention in recent years. To date, most recent graph embedding methods are evaluated on social and information networks and are not comprehensively studied on biomedical networks under systematic experiments and analyses. On the other hand, for a variety of biomedical network analysis tasks, traditional techniques such as matrix factorization (which can be seen as a type of graph embedding methods) have shown promising results, and hence there is a need to systematically evaluate the more recent graph embedding methods (e.g. random walk-based and neural network-based) in terms of their usability and potential to further the state-of-the-art. RESULTS: We select 11 representative graph embedding methods and conduct a systematic comparison on 3 important biomedical link prediction tasks: drug-disease association (DDA) prediction, drug-drug interaction (DDI) prediction, protein-protein interaction (PPI) prediction; and 2 node classification tasks: medical term semantic type classification, protein function prediction. Our experimental results demonstrate that the recent graph embedding methods achieve promising results and deserve more attention in the future biomedical graph analysis. Compared with three state-of-the-art methods for DDAs, DDIs and protein function predictions, the recent graph embedding methods achieve competitive performance without using any biological features and the learned embeddings can be treated as complementary representations for the biological features. By summarizing the experimental results, we provide general guidelines for properly selecting graph embedding methods and setting their hyper-parameters for different biomedical tasks. AVAILABILITY AND IMPLEMENTATION: As part of our contributions in the paper, we develop an easy-to-use Python package with detailed instructions, BioNEV, available at: https://github.com/xiangyue9607/BioNEV, including all source code and datasets, to facilitate studying various graph embedding methods on biomedical tasks. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.
Assuntos
Redes Neurais de Computação , Software , Interações Medicamentosas , Proteínas , SemânticaRESUMO
A woman in her mid-50s, hesitant about general anaesthesia due to a difficult airway, opted for neuraxial anaesthesia for L4 laminectomy with pedicle screw fixation (L3-L5). Preoperatively, she received 150 µg buprenorphine and 1 mg midazolam. In lateral position, a T8-T9 epidural catheter was placed, followed by segmental spinal anaesthesia (2.5 mL 0.5% hyperbaric bupivacaine+30 µg clonidine) at T10-T11. Prone positioning was executed using standard techniques. During the 6-7 hours surgery, three 7 mL epidural top-ups (2% lignocaine epinephrine) were administered at 90 min intervals. Haemodynamics remained stable with 2.5 L crystalloids, 350 mL packed red cells and three ephedrine doses (6 mg each). Sedation included 150 µg buprenorphine and two 1 mg midazolam doses. Postoperatively, she received epidural 0.25% bupivacaine for 2 days, systemic analgesics and was discharged on the sixth day.
Assuntos
Raquianestesia , Buprenorfina , Feminino , Humanos , Anestésicos Locais , Midazolam , Bupivacaína , Raquianestesia/métodosRESUMO
In the classical information theoretic framework, information "value" is proportional to how novel/surprising the information is. Recent work building on such notions claimed that false news spreads faster than truth online because false news is more novel and therefore surprising. However, another determinant of surprise, semantic meaning (e.g., information's consistency or inconsistency with prior beliefs), should also influence value and sharing. Examining sharing behavior on Twitter, we observed separate relations of novelty and belief consistency with sharing. Though surprise could not be assessed in those studies, belief consistency should relate to less surprise, suggesting the relevance of semantic meaning beyond novelty. In two controlled experiments, belief-consistent (vs. belief-inconsistent) information was shared more despite consistent information being the least surprising. Manipulated novelty did not predict sharing or surprise. Thus, classical information theoretic predictions regarding perceived value and sharing would benefit from considering semantic meaning in contexts where people hold pre-existing beliefs.
RESUMO
MOTIVATION: Inferring the underlying regulatory pathways within a gene interaction network is a fundamental problem in Systems Biology to help understand the complex interactions and the regulation and flow of information within a system-of-interest. Given a weighted gene network and a gene in this network, the goal of an inference algorithm is to identify the potential regulatory pathways passing through this gene. RESULTS: In a departure from previous approaches that largely rely on the random walk model, we propose a novel single-source k-shortest paths based algorithm to address this inference problem. An important element of our approach is to explicitly account for and enhance the diversity of paths discovered by our algorithm. The intuition here is that diversity in paths can help enrich different functions and thereby better position one to understand the underlying system-of-interest. Results on the yeast gene network demonstrate the utility of the proposed approach over extant state-of-the-art inference algorithms. Beyond utility, our algorithm achieves a significant speedup over these baselines. AVAILABILITY: All data and codes are freely available upon request.
Assuntos
Algoritmos , Biologia Computacional/métodos , Redes Reguladoras de Genes , Biologia de Sistemas/métodos , Modelos Teóricos , Saccharomyces cerevisiae/genéticaRESUMO
MOTIVATION: In recent years, Markov clustering (MCL) has emerged as an effective algorithm for clustering biological networks-for instance clustering protein-protein interaction (PPI) networks to identify functional modules. However, a limitation of MCL and its variants (e.g. regularized MCL) is that it only supports hard clustering often leading to an impedance mismatch given that there is often a significant overlap of proteins across functional modules. RESULTS: In this article, we seek to redress this limitation. We propose a soft variation of Regularized MCL (R-MCL) based on the idea of iteratively (re-)executing R-MCL while ensuring that multiple executions do not always converge to the same clustering result thus allowing for highly overlapped clusters. The resulting algorithm, denoted soft regularized Markov clustering, is shown to outperform a range of extant state-of-the-art approaches in terms of accuracy of identifying functional modules on three real PPI networks. AVAILABILITY: All data and codes are freely available upon request. CONTACT: srini@cse.ohio-state.edu SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.
Assuntos
Algoritmos , Mapeamento de Interação de Proteínas/métodos , Análise por Conglomerados , Cadeias de Markov , Proteínas de Saccharomyces cerevisiae/metabolismoRESUMO
Background and Aims: Recommendations on paediatric single-injection local anaesthetic (LA) dosing for peripheral nerve blocks (PNBs) are based on the children's weight and limited by weight-based toxicity concerns. In this study, we assessed the extent of circumferential spread and block characteristics following the injection of an age-based volume (age in years = LA volume) of 0.25% bupivacaine following popliteal sciatic nerve block (PSNB). Methods: Thirty children aged between 2 and 12 years with the American Society of Anesthesiologists (ASA) physical status I and II and undergoing foot and ankle surgical procedures were given single-injection ultrasound-guided subparaneural PSNB using 0.25% bupivacaine at age-based LA volume after the administration of anaesthesia. The circumferential pattern of LA spread (primary objective) was assessed along the nerve (both cephalad and caudal) using ultrasound from the point of administration and the block characteristics in terms of duration of sensory block. Results: The mean [standard deviation (SD)] cephalic circumferential LA spread distance was 2.52 (0.68) [95% confidence interval (CI): 2.27-2.76] cm. The mean (SD) caudal circumferential LA spread distance was 2.27 (0.48) [95% CI: 2.09-2.44] cm. The mean (SD) duration of the sensory block was 9.03 (0.97) [95% CI: 8.67-9.38] h. Conclusion: The age-based LA volume of bupivacaine for ultrasound-guided PSNB resulted in a longitudinal circumferential spread of around 4.7 cm (adding both cephalic and caudal spread) and provided adequate analgesia for nine postoperative hours.
RESUMO
BACKGROUND: Advances in high-throughput technology has led to an increased amount of available data on protein-protein interaction (PPI) data. Detecting and extracting functional modules that are common across multiple networks is an important step towards understanding the role of functional modules and how they have evolved across species. A global protein-protein interaction network alignment algorithm attempts to find such functional orthologs across multiple networks. RESULTS: In this article, we propose a scalable global network alignment algorithm based on clustering methods and graph matching techniques in order to detect conserved interactions while simultaneously attempting to maximize the sequence similarity of nodes involved in the alignment. We present an algorithm for multiple alignments, in which several PPI networks are aligned. We empirically evaluated our algorithm on three real biological datasets with 6 different species and found that our approach offers a significant benefit both in terms of quality as well as speed over the current state-of-the-art algorithms. CONCLUSION: Computational experiments on the real datasets demonstrate that our multiple network alignment algorithm is a more efficient and effective algorithm than the state-of-the-art algorithm, IsoRankN. From a qualitative standpoint, our approach also offers a significant advantage over IsoRankN for the multiple network alignment problem.
Assuntos
Algoritmos , Biologia Computacional/métodos , Mapas de Interação de Proteínas , Sequência de Aminoácidos , Animais , Análise por Conglomerados , Humanos , Proteínas/química , Proteínas/genética , Alinhamento de SequênciaRESUMO
Background and Aims: The efficacy of bilateral nasociliary and maxillary nerve blocks combined with general anaesthesia on intraoperative opioids consumption, emergence and recovery outcomes in adult patients is not well established. We conducted this study to test the hypothesis that the above blocks, combined with general anaesthesia, decrease the intraoperative opioid consumption following nasal surgery. Methods: In this prospective, double-blinded, randomised controlled study, 51 adult patients undergoing elective nasal surgery under general anaesthesia were randomised into one of two groups. Group A (n = 26) received bilateral nasociliary and maxillary nerve blocks with 12 mL of equal volumes of 0.5% bupivacaine and 2% lignocaine after induction of general anaesthesia. Group B (n = 25) did not receive any block (control group). The primary endpoint was the total intraoperative dose of fentanyl consumed. The secondary endpoints were the grade of cough, emergence agitation, the grade of post-operative nausea and vomiting, time to the first analgesia and time to post-anaesthesia care unit discharge. Results: The mean total intraoperative fentanyl dose (µg) was significantly lower in group A than in group B (2.31 ± 11.76 vs. 41.20 ± 31.00, P = 0.00). The incidence of emergence agitation was lower in group A than group B (11.5% vs. 88%, P = 0.00). The time to the first analgesia was significantly longer in group A than group B (543.27 vs. 199.84 min, P = 0.017). Conclusion: The pre-emptive administration of bilateral nasociliary and maxillary nerve block for nasal surgery is an effective technique for reducing the intraoperative dose of fentanyl and emergence agitation.
RESUMO
Background and Aims: The safety of conventional regional nerve block techniques in patients with established neuropathies, such as diabetic peripheral neuropathy (DPN), is still unclear. We designed this prospective dose finding study to identify the minimum effective local anaesthetic volume of 0.5% bupivacaine for ultrasound-guided subparaneural popliteal sciatic nerve block in 90% of DPN patients undergoing below-knee surgery (MELAV90). Methods: Fifty-three patients with diabetic peripheral neuropathy and scheduled for below knee surgical procedure received popliteal sciatic nerve block under ultrasound guidance. The initial local anaesthetic volume used was 10 ml of 0.5% bupivacaine. The subsequent local anaesthetic volume allocation was based on biased-coin-design. Accordingly, the local anaesthetic volume given to each subject was based on the block outcome of the previous patient. The study included patients prospectively until 45 successful blocks were obtained. The primary measurement was the minimum effective local anaesthetic volume resulting in a successful subparaneural popliteal sciatic nerve block in 90% of DPN patients. The MELAV90 was calculated using isotonic regression and a 95% confidence interval bootstrapping method. Results: The study included 53 patients to obtain 45 successful blocks. The MELAV90 of 0.5% bupivacaine was obtained at 5.85 ml (95% confidence interval, 5.72 to 6.22 ml). Eight patients needed supplemental anaesthesia to complete the surgery. No other complications were noted. Conclusion: For patients with diabetic peripheral neuropathy undergoing below-knee surgery, the MELAV90 of 0.5% bupivacaine for subparaneural popliteal sciatic nerve to achieve surgical anaesthesia was 5.85 ml.
RESUMO
Human mobility analysis plays a crucial role in urban analysis, city planning, epidemic modeling, and even understanding neighborhood effects on individuals' health. Often, these studies model human mobility in the form of co-location networks. We have recently seen the tremendous success of network representation learning models on several machine learning tasks on graphs. To the best of our knowledge, limited attention has been paid to identifying communities using network representation learning methods specifically for co-location networks. We attempt to address this problem and study user mobility behavior through the communities identified with latent node representations. Specifically, we select several diverse network representation learning models to identify communities from a real-world co-location network. We include both general-purpose representation models that make no assumptions on network modality as well as approaches designed specifically for human mobility analysis. We evaluate these different methods on data collected in the Adolescent Health and Development in Context (AHDC) study. Our experimental analysis reveals that a recently proposed method (LocationTrails) offers a competitive advantage over other methods with respect to its ability to represent and reflect community assignment that is consistent with extant findings regarding neighborhood racial and socio-economic differences in mobility patterns. We also compare the learned activity profiles of individuals by factoring in their residential neighborhoods. Our analysis reveals a significant contrast in the activity profiles of individuals residing in white-dominated vs. black-dominated neighborhoods and advantaged vs. disadvantaged neighborhoods in a major metropolitan city of United States. We provide a clear rationale for this contrastive pattern through insights from the sociological literature.
RESUMO
Many real-world applications deal with data that have an underlying graph structure associated with it. To perform downstream analysis on such data, it is crucial to capture relational information of nodes over their expanded neighborhood efficiently. Herein, we focus on the problem of Collective Classification (CC) for assigning labels to unlabeled nodes. Most deep learning models for CC heavily rely on differentiable variants of Weisfeiler-Lehman (WL) kernels. However, due to current computing architectures' limitations, WL kernels and their differentiable variants are limited in their ability to capture useful relational information only over a small expanded neighborhood of a node. To address this concern, we propose the framework, I-HOP, that couples differentiable kernels with an iterative inference mechanism to scale to larger neighborhoods. I-HOP scales differentiable graph kernels to capture and summarize information from a larger neighborhood in each iteration by leveraging a historical neighborhood summary obtained in the previous iteration. This recursive nature of I-HOP provides an exponential reduction in time and space complexity over straightforward differentiable graph kernels. Additionally, we point out a limitation of WL kernels where the node's original information is decayed exponentially with an increase in neighborhood size and provide a solution to address it. Finally, extensive evaluation across 11 datasets showcases the improved results and robustness of our proposed iterative framework, I-HOP.
RESUMO
The study of human cognition and the study of artificial intelligence (AI) have a symbiotic relationship, with advancements in one field often informing or creating new work in the other. Human cognition has many capabilities modern AI systems cannot compete with. One such capability is the detection, identification, and resolution of knowledge gaps (KGs). Using these capabilities as inspiration, we examine how to incorporate detection, identification, and resolution of KGs in artificial agents. We present a paradigm that enables research on the understanding of KGs for visual-linguistic communication. We leverage and enhance and existing KG taxonomy to identify possible KGs that can occur for visual question answer (VQA) tasks and use these findings to develop a classifier to identify questions that could be engineered to contain specific KG types for other VQA datasets. Additionally, we examine the performance of different VQA models through the lens of KGs.
Assuntos
Inteligência Artificial , Cognição , HumanosRESUMO
Due to delay in reporting, the daily national and statewide COVID-19 incidence counts are often unreliable and need to be estimated from recent data. This process is known in economics as nowcasting. We describe in this paper a simple random forest statistical model for nowcasting the COVID-19 daily new infection counts based on historic data along with a set of simple covariates, such as the currently reported infection counts, day of the week, and time since first reporting. We apply the model to adjust the daily infection counts in Ohio, and show that the predictions from this simple data-driven method compare favorably both in quality and computational burden to those obtained from the state-of-the-art hierarchical Bayesian model employing a complex statistical algorithm. The interactive notebook for performing nowcasting is available online at https://tinyurl.com/simpleMLnowcasting.
Assuntos
COVID-19 , Teorema de Bayes , Humanos , Incidência , Aprendizado de Máquina , SARS-CoV-2RESUMO
As Named Entity Recognition (NER) has been essential in identifying critical elements of unstructured content, generic NER tools remain limited in recognizing entities specific to a domain, such as drug use and public health. For such high-impact areas, accurately capturing relevant entities at a more granular level is critical, as this information influences real-world processes. On the other hand, training NER models for a specific domain without handcrafted features requires an extensive amount of labeled data, which is expensive in human effort and time. In this study, we employ distant supervision utilizing a domain-specific ontology to reduce the need for human labor and train models incorporating domain-specific (e.g., drug use) external knowledge to recognize domain specific entities. We capture entities related the drug use and their trends in government epidemiology reports, with an improvement of 8% in F1-score.
Assuntos
Armazenamento e Recuperação da Informação , Nomes , Humanos , Processamento de Linguagem NaturalRESUMO
Adaptive optics imaging has enabled the enhanced in vivo retinal visualization of individual cone and rod photoreceptors. Effective analysis of such high-resolution, feature rich images requires automated, robust algorithms. This paper describes RC-UPerNet, a novel deep learning algorithm, for identifying both types of photoreceptors, and was evaluated on images from central and peripheral retina extending out to 30° from the fovea in the nasal and temporal directions. Precision, recall and Dice scores were 0.928, 0.917 and 0.922 respectively for cones, and 0.876, 0.867 and 0.870 for rods. Scores agree well with human graders and are better than previously reported AI-based approaches.
RESUMO
Recent work uses a Siamese Network, initialized with BioWordVec embeddings (distributed word embeddings), for predicting synonymy among biomedical terms to automate a part of the UMLS (Unified Medical Language System) Metathesaurus construction process. We evaluate the use of contextualized word embeddings extracted from nine different biomedical BERT-based models for synonymy prediction in the UMLS by replacing BioWordVec embeddings with embeddings extracted from each biomedical BERT model using different feature extraction methods. Surprisingly, we find that Siamese Networks initialized with BioWordVec embeddings still outperform the Siamese Networks initialized with embedding extracted from biomedical BERT model.
RESUMO
The Unified Medical Language System (UMLS) Metathesaurus construction process mainly relies on lexical algorithms and manual expert curation for integrating over 200 biomedical vocabularies. A lexical-based learning model (LexLM) was developed to predict synonymy among Metathesaurus terms and largely outperforms a rule-based approach (RBA) that approximates the current construction process. However, the LexLM has the potential for being improved further because it only uses lexical information from the source vocabularies, while the RBA also takes advantage of contextual information. We investigate the role of multiple types of contextual information available to the UMLS editors, namely source synonymy (SS), source semantic group (SG), and source hierarchical relations (HR), for the UMLS vocabulary alignment (UVA) problem. In this paper, we develop multiple variants of context-enriched learning models (ConLMs) by adding to the LexLM the types of contextual information listed above. We represent these context types in context-enriched knowledge graphs (ConKGs) with four variants ConSS, ConSG, ConHR, and ConAll. We train these ConKG embeddings using seven KG embedding techniques. We create the ConLMs by concatenating the ConKG embedding vectors with the word embedding vectors from the LexLM. We evaluate the performance of the ConLMs using the UVA generalization test datasets with hundreds of millions of pairs. Our extensive experiments show a significant performance improvement from the ConLMs over the LexLM, namely +5.0% in precision (93.75%), +0.69% in recall (93.23%), +2.88% in F1 (93.49%) for the best ConLM. Our experiments also show that the ConAll variant including the three context types takes more time, but does not always perform better than other variants with a single context type. Finally, our experiments show that the pairs of terms with high lexical similarity benefit most from adding contextual information, namely +6.56% in precision (94.97%), +2.13% in recall (93.23%), +4.35% in F1 (94.09%) for the best ConLM. The pairs with lower degrees of lexical similarity also show performance improvement with +0.85% in F1 (96%) for low similarity and +1.31% in F1 (96.34%) for no similarity. These results demonstrate the importance of using contextual information in the UVA problem.