Context Of Inference Research Articles

On the computational complexity of Dempster’s Rule of combination, a parallel computing approach

Multisensor data fusion with Dempster–Shafer (D–S) theory is beneficial for context inference without any advanced information. D–S theory includes reasoning based on Dempster’s rule of combination of degrees of belief based on different pieces of evidence. Although, the computational complexity of Dempster’s rule of combination is enormous. Any change in the number of pieces of evidence or hypotheses causes a lot of additional computations. Dempster’s rule of combination is shown to be #P-complete. The combination rule of k frames of evidence with nk elements, such that (k∈N|k≥2) has a time complexity of O(2N), where N=∑knk. In this paper, we propose a parallel computing approach for Dempster’s rule of combination based on the concept of conquer and divide algorithms. The proportion of task benefiting from improvement is p=1−2k, hence k2 of theoretical maximum speed-up according to Amdahl’s law. We have tested our algorithm in different experimental settings and observed that the new parallel computing approach has not only achieved the best results in CPU version, but has also outperformed GPU version using Thrust CUDA in almost scenarios of the experiment.

Filosofia da Lógica Empiricamente Informada: uma ilustração via metafísica da lógica

Dados da psicologia são relevantes na abordagem de alguns problemas em filosofia da lógica mesmo que não se assuma, de antemão, uma posição psico-logista. Para ilustrar um método de abordar problemas na filosofia da lógica que faz largo uso de dados da psicologia, neste artigo considero a questão sobre a metafísica da lógica—qual o objeto de estudo da lógica?—à luz de resultados da psicologia do pensamento e da psicologia do desenvolvimento. Estes resultados nos permitem concluir que sistemas lógicos não são meramente descritivos de aspectos gerais da linguagem natural, do mundo ou da racionalidade humana. Os dados são consistentes com a hipótese de que sis-temas lógicos dedutivos sejam conjuntos de regras de caráter primordialmente não-descritivo, criadas mediante reflexão ativa de filósofos, matemáticos e lógicos, visando garantir inferências dedutivas em certos contextos.

Similative plurality and the nature of alternatives

This paper investigates the properties of similative plurals, focusing on m-reduplication in Persian and -toka and -tari in Japanese. Although these expressions are associated with what I refer to as a non-homogeneous plural inference in upward-entailing contexts, I demonstrate that this inference is not an entailment of sentences with these morphemes, but is merely implicated, much like the multiplicity condition associated with English bare plurals (Krifka 2004; Spector 2007; Zweig (2009); de Swart & Farkas 2010). I propose an analysis of similative plurals as mereological mixtures of a set with a set of contextually similar objects, and derive the non-homogeneous plural reading via scalar implicature. I demonstrate that deriving this implicature requires both the calculation of implicature at a subsentential level (Chierchia 2004; Chierchia 2006; Zweig 2009) and appeal to an abstract alternative (Buccola et al. 2020; Charlow 2019). This latter point provides a challenge for theories of alternative generation based on structural replacements and deletions (Katzir 2007). EARLY ACCESS

Zero-Shot Recognition Enhancement by Distance-Weighted Contextual Inference

Zero-shot recognition (ZSR) aims to perform visual classification by category in the absence of training samples. The focus in most traditional ZSR models is using semantic knowledge about familiar categories to represent unfamiliar categories with only the visual appearance of an unseen object. In this research, we consider not only visual information but context to enhance the classifier’s cognitive ability in a multi-object scene. We propose a novel method, contextual inference, that uses external resources such as knowledge graphs and semantic embedding spaces to obtain similarity measures between an unseen object and its surrounding objects. Using the intuition that close contexts involve more related associations than distant ones, distance weighting is applied to each piece of surrounding information with a newly defined distance calculation formula. We integrated contextual inference into traditional ZSR models to calibrate their visual predictions, and performed extensive experiments on two different datasets for comparative evaluations. The experimental results demonstrate the effectiveness of our method through significant enhancements in performance.

Cenogram analyses as habitat indicators for Paleogene–Neogene mammalian communities across the globe, with an emphasis on the early Eocene Cambay Shale mammalian community from India

The present study is an attempt to utilize cenogram methodology (both qualitative and quantitative) to consider mammalian communities from five early to late Eocene localities across the globe (i.e., Polecat Bench, Bighorn Basin, North America; Abbey Wood, Blackheath Formation, UK; Cambay Shale, Cambay Basin, India; Wutu Formation, Wutu Basin, China; Pondaung Formation, Myanmar) so as to provide a comparative palaeohabitat framework. It is also a first attempt to examine the palaeohabitat of an extinct mammalian community (i.e., from Cambay Shale) in India utilizing the cenogram approach. In addition, seven extinct middle-Miocene communities (Laogou, Linxia Basin, China; Estacion Imperial, Spain; Paseo de las Acacias, Spain; Arroyo del Olivar-Puente de Vallecas, Spain; Somo-saguas, Spain; Paracuellos 5, Spain; Paracuellos 3, Spain) have also been considered, in order to provide a global perspective to the climatic inferences in a temporal context. The majority of statistical calculations for Paleogene communities expose forested and humid conditions, excluding the Cambay Shale mammalian community of India. A hidden diversity within the medium to large body-size category of mammals (disguising the mammal biodiversity expected in tropical forested habitats) from Cambay Shale (western India) is a plausible cause of digression in the results. This is reflected in the histograms showing relationships between proportions of mammal species in various body-mass categories. Furthermore, the results show that Neogene mammalian communities were sustained in comparatively open habitats. Differences between occidental European and Asian localities in the canopy and humidity of the Neogene environments are also reflected in our analyses.

Auto-Encoding and Distilling Scene Graphs for Image Captioning.

We propose scene graph auto-encoder (SGAE) that incorporates the language inductive bias into the encoder-decoder image captioning framework for more human-like captions. Intuitively, we humans use the inductive bias to compose collocations and contextual inferences in discourse. For example, when we see the relation "a person on a bike", it is natural to replace "on" with "ride" and infer "a person riding a bike on a road" even when the "road" is not evident. Therefore, exploiting such bias as a language prior is expected to help the conventional encoder-decoder models reason as we humans and generate more descriptive captions. Specifically, we use the scene graph-a directed graph ( G) where an object node is connected by adjective nodes and relationship nodes-to represent the complex structural layout of both image ( I) and sentence ( S). In the language domain, we use SGAE to learn a dictionary set ( D) that helps reconstruct sentences in the S→ GS → D → S auto-encoding pipeline, where D encodes the desired language prior and the decoder learns to caption from such a prior; in the vision-language domain, we share D in the I→ GI → D → S pipeline and distill the knowledge of the language decoder of the auto-encoder to that of the encoder-decoder based image captioner to transfer the language inductive bias. In this way, the shared D provides hidden embeddings about descriptive collocations to the encoder-decoder and the distillation strategy teaches the encoder-decoder to transform these embeddings to human-like captions as the auto-encoder. Thanks to the scene graph representation, the shared dictionary set, and the Knowledge Distillation strategy, the inductive bias is transferred across domains in principle. We validate the effectiveness of SGAE on the challenging MS-COCO image captioning benchmark, where our SGAE-based single-model achieves a new state-of-the-art 129.6 CIDEr-D on the Karpathy split, and a competitive 126.6 CIDEr-D (c40) on the official server, which is even comparable to other ensemble models. Furthermore, we validate the transferability of SGAE on two more challenging settings: transferring inductive bias from other language corpora and unpaired image captioning. Once again, the results of both settings confirm the superiority of SGAE. The code is released in https://github.com/yangxuntu/SGAE.

Accuracy of the clinical indicators of ineffective health management in celiac people.

to analyze the accuracy of clinical indicators of "Ineffective health management" in celiac patients and to verify associations between sociodemographic characteristics and clinical indicators. a cross-sectional study, conducted from May to September 2017, with 83 celiac patients, through an interview. Accuracy measures were defined by latent class model. there was a prevalence of "Ineffective health management" of 55.69%. "Failure to take action to reduce risk factor" and "Failure to include treatment regimen in daily living" better predict this diagnosis. Paid occupation reduces the chance of the presence of "Difficulty with prescribed regimen". Participation in support association reduces the chance of the presence of "Difficulty with prescribed regimen", "Ineffective choices in daily living for meeting health goal" and "Failure to take action to reduce risk factor". accurate clinical indicators identification assists clinical reasoning for diagnostic inference in specific health contexts.

Reconstruction of Femur Length Using Epiphyseal and Diaphyseal Diameters in Contemporary Egyptian Sample, with Application to Ancient Egyptians.

Inferences in bioarchaeology and forensic contexts require mathematical stature estimation using long bone lengths. This study aimed to identify predictors of femur length (FL) from epiphyseal and diaphyseal width measurements that are not bound to assumptions of sex or laterality. To compute linear regression models, both standard and new measurements around the diaphyseal dominant nutrient foramina (NF) were collected on modern femora (n = 64) from the unidentified skeletal collection housed at Alexandria University. Four equations were then validated on an ancient Egyptian sample (n = 73) from the Goldman Osteometric Data Set to evaluate the effect of sex subdivision on the accuracy of FL and indirect stature estimations using Raxter's formulas. Most of models reflected significant positive association (r > 0.60) between width variables and FL. Oddly, the distance from the proximal end to the NF correlated weakly with FL (r = 0.34). The stepwise selected equations preferred measurements around NF to midshaft, with the anteroposterior diameter included in the proximal fragment model (r = 0.77), and circumference in diaphyseal fragment model (r = 0.62). Tested equations performed consistently on the ancient Egyptian sample. Measurements from the femoral proximal fragment are more reliable predictors than those from the distal fragment, with the exception of femur neck diameter. However, distal epicondylar breadth is a better predictor of FL in females than in males. Indirect stature estimation showed a reasonable degree of accuracy in both sexes. These models can be applied successfully in contemporary and ancient Egyptian fragmentary remains; however, due to larger size of femora from the Old Kingdom sample, they would be most applicable to individuals from later dynasties.

Vegetation Land Use/Land Cover Extraction From High-Resolution Satellite Images Based on Adaptive Context Inference

In this paper, automatic extraction of multi-context and multi-scale land use/land cover vegetation from high-resolution remote sensing images is tackled, aiming to solve typical challenges in classifying remote sensing images at a pixel level. To solve small inter-class differences and large intra-class differences between the vegetation and background, we introduce a vegetation-feature-sensitive focus perception (FP) module. Considering the intrinsic properties of vegetation objects, we established an adaptive context inference (ACI) model under a supervised setting that includes a context model to represent relationships between a center pixel and its neighbors under semantic constraints, as well as the spatial structures of vegetation features. Comparative experiments on the ZY-3 and Gaofen Image Dataset (GID) datasets demonstrate the effectiveness of our proposed automatic vegetation extraction model against the baseline Deeplab v3+ model. Taking precision, kappa coefficient, mean intersection over union (miou), precision rate, and F1-score as the evaluation indexes, the results showed an improvement in the precision by at least 1.44% and miou by 2.47%, over the baseline Deeplab v3+ model. In addition, the ACI module improved the precision and miou by 2% and 3.88%, and the FP module improved the precision and miou by 1.13% and 1.65%. These results and statistics of these comprehensive experiments illustrated that our adaptive and effective vegetation extraction model could satisfy different requirements of land use/land cover mapping applications.

On Presupposition Projection with Trivalent Connectives

A basic puzzle about presuppositions concerns their projection from propositional constructions. This problem has regained much attention in the last decade since many of its prominent accounts, including variants of the trivalent Strong Kleene connectives, suffer from the so-called *proviso problem*.This paper argues that basic insights of the Strong Kleene system can be used without invoking the proviso problem. It is shown that the notion of *determinant value* that underlies the definition of the Strong Kleene connectives leads to a natural generalization of the filtering conditions proposed in Karttunen's article ``Presuppositions of compound sentences'' (LI, 1973). Incorporating this generalized condition into an incremental projection algorithm avoids the proviso problem as well as the derivation of conditional presuppositions. It is argued that the same effects that were previously modelled using conditional presuppositions may be viewed as effects of presupposition suspension and contextual inference on presupposition projection.

A flexible and generalizable model of online latent-state learning.

Many models of classical conditioning fail to describe important phenomena, notably the rapid return of fear after extinction. To address this shortfall, evidence converged on the idea that learning agents rely on latent-state inferences, i.e. an ability to index disparate associations from cues to rewards (or penalties) and infer which index (i.e. latent state) is presently active. Our goal was to develop a model of latent-state inferences that uses latent states to predict rewards from cues efficiently and that can describe behavior in a diverse set of experiments. The resulting model combines a Rescorla-Wagner rule, for which updates to associations are proportional to prediction error, with an approximate Bayesian rule, for which beliefs in latent states are proportional to prior beliefs and an approximate likelihood based on current associations. In simulation, we demonstrate the model’s ability to reproduce learning effects both famously explained and not explained by the Rescorla-Wagner model, including rapid return of fear after extinction, the Hall-Pearce effect, partial reinforcement extinction effect, backwards blocking, and memory modification. Lastly, we derive our model as an online algorithm to maximum likelihood estimation, demonstrating it is an efficient approach to outcome prediction. Establishing such a framework is a key step towards quantifying normative and pathological ranges of latent-state inferences in various contexts.

Comprehension and Inference: Relationships Between Oral and Written Modalities in Good and Poor Comprehenders During Adolescence

Purpose We investigated the relationships between text reading comprehension and oral idiom comprehension in adolescents. We also examined the more specific relationships between inference in text comprehension and inference in idiom comprehension. Method We selected participants from an initial sample of 140 students aged 13-15 years to form 2 groups, according to their decoding and reading comprehension abilities: 1 group of good comprehenders/good decoders (n = 49) and 1 group of less skilled comprehenders but with adequate decoding skills (n = 20). The reading comprehension task comprised both literal and inferential (text-based and knowledge-based) questions. These 2 groups were then compared on an idiom comprehension task. In this task, idioms were presented orally, and students were placed in a situation that simulated a real-life oral interaction. The idioms were novel for the students (translated from a foreign language), either transparent or opaque, and presented either with a supportive context or without any context. Results Good reading comprehenders outperformed less skilled ones on the idiom task. Both groups benefited from the supportive context, especially the good comprehenders. Knowledge-based inferences in written text comprehension were related to contextual inferences for opaque idioms, while semantic inferences for transparent idioms were related to literal text comprehension, but not to text-connecting inferences. Conclusion These results are discussed both theoretically, in terms of cross-modal comprehension processes, and practically, in terms of implications for remediation.

Fuzzy contextual inference system for medical image enhancement

PurposeIn order to facilitate early detection and further diagnosis, enhancing contrast and preserving natural characteristics are the fundamental requirement. In MR imaging, the poor quality image specifically the low contrast image may deliver inadequate data for the visual interpretation of affected portions. MethodsFuzzy Contextual Dissimilarity Adaptive Histogram Equalization (FCDAHE) is employed to enhance the contrast of medical images. The proposed method consists of two modules. In the first module, fuzzy dissimilarity histogram is formulated by finding fuzzy neighbourhood dissimilarity for every pixel using fuzzy rules. Then, the fuzzy dissimilarity clip limit is computed based on a fuzzy inference system. Contrast (C) and discrete entropy (E) are the two parameters used by the Fuzzy Inference System (FIS) to achieve fuzzy dissimilarity clip limit. Then, the enhanced output is obtained by equalizing the fuzzy dissimilarity adaptive histogram. In the second module, Contextual Intensity Transformation (CIT) is applied to fuzzy dissimilarity adaptive histogram equalized output to get final enhanced images. ResultsExperiments are conducted and tested on a wide variety of MR images extracted from BRATS 2015 database and Magnetic Resonance – Technology Information Portal to evaluate the performance of the proposed method both qualitatively and quantitatively. Extensive quantitative measures show that the proposed technique yields better performance in terms of PSNR, entropy, contrast ratio and EME when compared to state of art enhancement methods. ConclusionThe proposed FCDAHE algorithm not only improves contrast but also preserves natural characteristics. The proposed method offers a better scope for disease analysis and diagnosis.

Self-adaptation modeling for service evolution on the Internet of Things (IoT)

One of the most important concepts of the internet of things is related to software services, where the system must be able to provide real-time data collected from various environments. So that software services represent a variety of physical and virtual real-world objects that can grow very fast. This condition is related to the ability of self-adaptation software services. However, the existing model of self-adaptation is generally not paying attention to the requirements for service evolution. The objective of this paper is to introduce self-adaptation modeling techniques which consist of, first, domain modeling of the internet of things to represent real-world context; second, developing an inference engine for context inference. As a form of evaluation, this model is applied to the patient monitoring system that will relate to the concept of self-adaptation of various systems, devices, actors, and environment. The case study results show the system’s ability to anticipate changes in context and growth needs.

Beyond Prediction: A Framework for Inference With Variational Approximations in Mixture Models

Variational inference is a popular method for estimating model parameters and conditional distributions in hierarchical and mixed models, which arise frequently in many settings in the health, social, and biological sciences. Variational inference in a frequentist context works by approximating intractable conditional distributions with a tractable family and optimizing the resulting lower bound on the log-likelihood. The variational objective function is typically less computationally intensive to optimize than the true likelihood, enabling scientists to fit rich models even with extremely large datasets. Despite widespread use, little is known about the general theoretical properties of estimators arising from variational approximations to the log-likelihood, which hinders their use in inferential statistics. In this article, we connect such estimators to profile M-estimation, which enables us to provide regularity conditions for consistency and asymptotic normality of variational estimators. Our theory also motivates three methodological improvements to variational inference: estimation of the asymptotic model-robust covariance matrix, a one-step correction that improves estimator efficiency, and an empirical assessment of consistency. We evaluate the proposed results using simulation studies and data on marijuana use from the National Longitudinal Study of Youth. Supplementary materials for this article are available online.

Literature reading modulates pronoun resolution in counterfactual world: Evidence from event-related potentials.

Counterfactuals are contrary-to-fact statements that are widely used in daily life to convey thoughts about what might have been. Different from fact-based processing, successful counterfactual comprehension requires readers to keep in mind both suppositional information and presupposed fact. Using event-related potentials, the present study investigates how the process of establishing a coreferential relation (i.e., pronoun resolution) is influenced by counterfactual context, and whether it will be modulated by individual difference in literature reading. We compared the P600 (a positive-going deflection, which often reaches its peak around 600 milliseconds after presentation of the stimulus) effects elicited by gender-mismatched pronouns in three conditionals (causal vs. hypothetical vs. counterfactual) between two groups (literature exposure high- vs. low-level group). Results show that for low-level group, incongruent pronouns elicited robust P600 effects across all three conditionals, while for high-level group, the P600 effects were pronounced only in causal and hypothetical conditionals, but not in counterfactual conditionals. These findings suggest (a) different from causal and hypothetical conditionals, the dual meaning and pragmatic implications of counterfactuals may prompt people to go beyond here and now to elaborate their mental models and entertain alternative interpretations, and (b) substantial literature exposure would further enhance pragmatic inference of counterfactual context, leading high-level readers more inclined to elaborate discourse with possible alternative inferences, while leaving low-level readers habitually resort to more straightforward coreferential interpretation. (PsycINFO Database Record (c) 2019 APA, all rights reserved).

Inferential language processes across adolescence: A test with novel idioms.

Idiom comprehension relies on the ability to draw inferences from different cues in a communication situation. Adopting a developmental perspective, we investigated how this ability changes across adolescence. To this end, we designed a computerized system that allowed us to simulate a communication situation through short videos placing participants at the center of the interaction. Four groups of participants (11, 13, 15, and adult students) performed an idiomatic expression comprehension task, in which idiom familiarity was controlled. We manipulated the idioms' transparency (vs. opacity) and presentation (supportive narrative context inducing an idiomatic interpretation of the expression vs. nonsupportive narrative context). Analyses revealed an improvement in idiom comprehension in terms of contextual inferences but they failed to reveal any significant difference in terms of semantic inferences. This study yields fresh arguments in support of the notion that inferential ability based on context continues to develop beyond childhood.

Biased political reasoning and relational inferences in a small-group deliberative context

Theorists of deliberative democracy envision a citizenry engaged in collective reasoning about public issues on the merits of rival arguments. Partisanship undermines this ideal when it causes people to discount counter-attitudinal arguments, independent of their quality. Empirical deliberative theory lacks an account for what mechanism mitigates this bias in small-group settings. To close that theoretical gap, this study draws on Relational Framing Theory and identifies a relational component of the reasoning process. Participants rated the relevance of dominance/submission and affiliation/disaffiliation relational frames after a small-group deliberation. This perception influenced participants’ decisions to endorse arguments as legitimate public reasons. Implications for deliberative theory, research, and practice are discussed.

Prior Precision Modulates the Minimization of Auditory Prediction Error

The predictive coding model of perception proposes that successful representation of the perceptual world depends upon canceling out the discrepancy between prediction and sensory input (i.e., prediction error). Recent studies further suggest a distinction to be made between prediction error triggered by non-predicted stimuli of different prior precision (i.e., inverse variance). However, it is not fully understood how prediction error with different precision levels is minimized in the predictive process. Here, we conducted a magnetoencephalography (MEG) experiment which orthogonally manipulated prime-probe relation (for contextual precision) and stimulus repetition (for perceptual learning which decreases prediction error). We presented participants with cycles of tone quartets which consisted of three prime tones and one probe tone of randomly selected frequencies. Within each cycle, the three prime tones remained identical while the probe tones changed once at some point (e.g., from repetition of 123X to repetition of 123Y). Therefore, the repetition of probe tones can reveal the development of perceptual inferences in low and high precision contexts depending on their position within the cycle. We found that the two conditions resemble each other in terms of N1m modulation (as both were associated with N1m suppression) but differ in terms of N2m modulation. While repeated probe tones in low precision context did not exhibit any modulatory effect, repeated probe tones in high precision context elicited a suppression and rebound of the N2m source power. The differentiation suggested that the minimization of prediction error in low and high precision contexts likely involves distinct mechanisms.

Belief of agency changes dynamics in sensorimotor networks

Controlling an event through one’s own action usually induces a sense of agency, a feeling that arises when an expected outcome matches the intention. The neural correlates of this feeling remain controversial however, as experimental manipulation of the action-outcome chain often introduces mismatch or prediction errors that strongly correlate with the sense of agency. Here, we took a different approach and manipulated the causal belief (self-attribution vs. computer-attribution) by external cues during matched visuo-motor tapping conditions. With magneto-encephalography, we studied the sense of agency from a network perspective, investigating in source space the modulation of local population activity and changes in functional connectivity with motor cortex. Our results show that during the belief of agency primary motor cortex (M1) shows stronger functional connectivity (mediated by the beta band) to inferior parietal lobe and right middle temporal gyrus (MTG). Furthermore, the local feed-forward activity (gamma band power) in extrastriate body area and MTG disappears with that belief. After changes in action context, left M1 shows stronger connectivity in the alpha band with right premotor cortex and left insular-temporal cortex a network that might support active inference in social action context. Finally, a better tapping performance in this rhythmic task was related to alpha power modulations in the bilateral cerebellum and bilateral fusiform body-area, with power suppression during a more precise performance. These findings highlight the role of multiple networks supporting the sense of agency by changing their relative contribution for different causal beliefs.