Continuous patient state attention model for addressing irregularity in electronic health records.

BACKGROUND: Irregular time series (ITS) are common in healthcare as patient data is recorded in an electronic health record (EHR) system as per clinical guidelines/requirements but not for research and depends on a patient's health status. Due to irregularity, it is challenging to develop machine learning techniques to uncover vast intelligence hidden in EHR big data, without losing performance on downstream patient outcome prediction tasks. METHODS: In this paper, we propose Perceiver, a cross-attention-based transformer variant that is computationally efficient and can handle long sequences of time series in healthcare. We further develop continuous patient state attention models, using Perceiver and transformer to deal with ITS in EHR. The continuous patient state models utilise neural ordinary differential equations to learn patient health dynamics, i.e., patient health trajectory from observed irregular time steps, which enables them to sample patient state at any time. RESULTS: The proposed models' performance on in-hospital mortality prediction task on PhysioNet-2012 challenge and MIMIC-III datasets is examined. Perceiver model either outperforms or performs at par with baselines, and reduces computations by about nine times when compared to the transformer model, with no significant loss of performance. Experiments to examine irregularity in healthcare reveal that continuous patient state models outperform baselines. Moreover, the predictive uncertainty of the model is used to refer extremely uncertain cases to clinicians, which enhances the model's performance. Code is publicly available and verified at https://codeocean.com/capsule/4587224 . CONCLUSIONS: Perceiver presents a computationally efficient potential alternative for processing long sequences of time series in healthcare, and the continuous patient state attention models outperform the traditional and advanced techniques to handle irregularity in the time series. Moreover, the predictive uncertainty of the model helps in the development of transparent and trustworthy systems, which can be utilised as per the availability of clinicians.

Individual differences in dissimilation: Do some people make more distinctions among targets' personalities than others?

OBJECTIVE: People differ in how positively they tend to see others' traits, but people might also differ in how strongly they apply their perceptual styles. In two studies (Ns = 355, 303), the current research explores individual differences in how variable people's first impressions are across targets (i.e., within-person variability), how and why these differences emerge, and who varies more in their judgments of others. METHOD: Participants described themselves on personality measures and rated 30 (Study 1) or 90 (Study 2) targets on Big Five traits. RESULTS: Using the extended Social Relations Model (eSRM), results suggest that within-person variability in impressions is consistent across trait ratings. People lower in extraversion, narcissism and self-esteem tended to make distinctions across targets' Big Five traits that were more consistent with other perceivers (sensitivity). Furthermore, some people more than others tended to consistently make unique distinctions among targets (differentiation), and preliminary evidence suggests these people might be higher in social anxiety and lower in self-esteem and emotional stability. CONCLUSION: Overall then, a more complete account of person perception should consider individual differences in how variable people's impressions are of others.

Predicting gene and protein expression levels from DNA and protein sequences with Perceiver.

BACKGROUND AND OBJECTIVE: The functions of an organism and its biological processes result from the expression of genes and proteins. Therefore quantifying and predicting mRNA and protein levels is a crucial aspect of scientific research. Concerning the prediction of mRNA levels, the available approaches use the sequence upstream and downstream of the Transcription Start Site (TSS) as input to neural networks. The State-of-the-art models (e.g., Xpresso and Basenjii) predict mRNA levels exploiting Convolutional (CNN) or Long Short Term Memory (LSTM) Networks. However, CNN prediction depends on convolutional kernel size, and LSTM suffers from capturing long-range dependencies in the sequence. Concerning the prediction of protein levels, as far as we know, there is no model for predicting protein levels by exploiting the gene or protein sequences. METHODS: Here, we exploit a new model type (called Perceiver) for mRNA and protein level prediction, exploiting a Transformer-based architecture with an attention module to attend to long-range interactions in the sequences. In addition, the Perceiver model overcomes the quadratic complexity of the standard Transformer architectures. This work's contributions are 1. DNAPerceiver model to predict mRNA levels from the sequence upstream and downstream of the TSS; 2. ProteinPerceiver model to predict protein levels from the protein sequence; 3. Protein&DNAPerceiver model to predict protein levels from TSS and protein sequences. RESULTS: The models are evaluated on cell lines, mice, glioblastoma, and lung cancer tissues. The results show the effectiveness of the Perceiver-type models in predicting mRNA and protein levels. CONCLUSIONS: This paper presents a Perceiver architecture for mRNA and protein level prediction. In the future, inserting regulatory and epigenetic information into the model could improve mRNA and protein level predictions. The source code is freely available at https://github.com/MatteoStefanini/DNAPerceiver.

Committing Fallacies and the Appearance Condition.

This appearance condition of fallacies refers to the phenomenon of weak arguments, or moves in argumentation, appearing to be okay when really they aren't. Not all theorists agree that the appearance condition should be part of the conception of fallacies but this essay explores some of the consequences of including it. In particular, the differences between committing a fallacy, causing a fallacy and observing a fallacy are identified. The remainder of the paper is given over to discussing possible causes of mistakenly perceiving weak argumentation moves as okay. Among these are argument caused misperception, perspective caused misperception, discursive environment caused misperception and perceiver caused misperception. The discussion aims to be sufficiently general so that it can accommodate different models and standards of argumentation that make a place for fallacies.

Seeing the Best or Worst in Others: A Measure of Generalized Other-Perceptions.

How positively or negatively people generally view others is key for understanding personality, social behavior, and psychopathology. Previous research has measured generalized other-perceptions by relying on either explicit self-reports or judgments made in group settings. With the current research, we overcome the limitations of these past approaches by introducing a novel measurement instrument for generalized other-perceptions: the Online-Tool for Assessing Perceiver Effects (O-TAPE). By assessing perceivers' first impressions of a standardized set of target people displayed in social network profiles or short video sequences, the O-TAPE captures individual differences in the positivity of other-perceptions. In Study 1 (n = 219), the instrument demonstrated good psychometric properties and correlations with related constructs. Study 2 (n = 142) replicated these findings and also showed that the O-TAPE predicted other-perceptions in a naturalistic group setting. Study 3 (n = 200) refined the nomological network of the construct and demonstrated that the O-TAPE is invulnerable to effects of social desirability.

Asymmetries in Mutual Understanding: People With Low Status, Power, and Self-Esteem Understand Better Than They Are Understood.

All too often, people who develop exceptionally astute insights into others remain mysterious to these others. Evidence for such asymmetric understanding comes from several independent domains. Striking asymmetries occur among those who differ in status and power, such that individuals with low status and power understand more than they are understood. We show that this effect extends to people who merely perceive that they have low status: individuals with low self-esteem. Whereas people with low self-esteem display insight into people with high self-esteem, people with high self-esteem fail to reciprocate. Conceptual analysis suggests that asymmetries in mutual understanding may be reduced by addressing deficits in information and motivation among perceivers. Nevertheless, several interventions have been unsuccessful, indicating that the path to symmetric understanding is a steep and thorny one. Further research is needed to develop strategies for fostering understanding of those who are most misunderstood: people with low self-esteem, low status, and low power.

The effects of face attractiveness on face memory depend on both age of perceiver and age of face.

Face attractiveness can influence memory for previously seen faces. This effect has been shown to differ for young and older perceivers. Two parallel studies examined the moderation of both the age of the face and the age of the perceiver on the relationship between facial attractiveness and face memory. Study 1 comprised 29 young and 31 older participants; Study 2 comprised 25 young and 24 older participants. In both studies, participants completed an incidental face encoding and a surprise old/new recognition test with young and older faces that varied in face attractiveness. Face attractiveness affected memory for young but not older faces. In addition, young but not older perceivers showed a linear effect of facial attractiveness on memory for young faces, while both young and older perceivers showed a quadratic effect on memory for young faces. These findings extend previous work by demonstrating that the effect of facial attractiveness on face memory is a function of both the age of the perceiver and the age of the face. Factors that could account for such moderations of face and perceiver age on the associations between face attractiveness and face memory are discussed (e.g. age differences in social goals and face similarity/distinctiveness).

The Effect of Gender on Perceptions of Credibility in Child Sexual Assault Cases: A Systematic Review.

Child sexual assault cases have one of the highest attrition rates throughout the justice system, with jurors' perceptions of the credibility of the child playing a critical role in the judicial process and resulting outcomes. One of the most prominent influences on credibility is gender, specifically victim and perceiver gender. This article reports the findings of a systematic search of the literature exploring the impact of these factors on perceptions of victim credibility. Results suggest that overall females tend to rate victim credibility higher than do males; however, this gender effect is minimized when the alleged perpetrator is not the biological parent of the victim. There is little support for a victim gender effect, such that potential jurors' ratings of victim credibility does not appear to be influenced by the gender of the victim. The review highlights the lack of consistent and comprehensive measurement of credibility. Legal implications are discussed.

Quantifying the Association of Self-Enhancement Bias With Self-Ratings of Personality and Life Satisfaction.

Kwan, John, Kenny, Bond, and Robins conceptualize self-enhancement as a favorable comparison of self-judgments with judgments of and by others. Applying a modified version of Kwan et al.'s approach to behavior observation data, we show that the resulting measure of self-enhancement bias is highly reliable, predicts self-ratings of intelligence as well as does actual intelligence, interacts with item desirability in predicting responses to questionnaire items, and also predicts general life satisfaction. Consistent with previous research, however, self-ratings of intelligence did not become more valid when controlling for self-enhancement bias. We also show that common personality scales like the Rosenberg Self-Esteem Scale reflect self-enhancement at least as strongly as do scales that were designed particularly for that purpose (i.e., "social desirability scales"). The relevance of these findings in regard to the validity and utility of social desirability scales is discussed.