Modelling the human lane-change execution behaviour through Multilayer Perceptrons and Convolutional Neural Networks

Driving is a highly complex task that involves the execution of multiple cognitive tasks belonging to different levels of abstraction. Traffic emerges from the interaction of a big number of agents implementing those behaviours, but until recent years, modelling it by the interaction of these agents in the so called micro-simulators was a nearly impossible task as their number grows. However, with the growing computing power it is possible to model increasingly large quantities of individual vehicles according to their individual behaviours. These models are usually composed of two sub-models for two well-defined tasks: car-following and lane-change. In the case of lane-change the literature proposes many different models, but few of them use Computational Intelligence (CI) techniques, and much less use personalization for reaching individual granularity. This study explores one of the two aspects of the lane-change called lane-change acceptance, where the driver performs or not a lane-change given his intention and the vehicle environment. We demonstrate how the lane-change acceptance of a specific driver can be learned from his lane change intention and surrounding environment in an urban scenario using CI techniques such as feed-forward Artificial Neural Network (ANN). We work with Multilayer Perceptron (MLP) and Convolutional Neural Networks (CNN) architectures. How they perform one against the other and how the different topologies affect both to the generalization of the problem and the learning process are studied.     

Enterprise credit risk evaluation based on neural network algorithm

To explore the enterprise credit risk evaluation, the application effect of several common neural network models in Chinese small and medium-sized enterprise data sets was compared and the optimal parameters for each model were determined. In addition, the classification accuracy and the applicability of the model were compared, and finally the common problem of optimization neural network algorithm based on population was solved: need to determine the dimensions in advance. The experimental results showed that the probabilistic neural network (PNN) had the minimum error rate and second types of errors, while the PNN model had the highest AUC value and was robust. To sum up, the algorithm makes some contributions to solve the financing problem of small and medium-sized enterprises in China.     

Attentional orienting and awareness: evidence from a discrimination task

We used several cue–target SOAs (100, 500, 1000 ms) and three different degrees of cue predictability (Non-predictive-50%, Predictive-75%, Counter-predictive-25%), to investigate the role of awareness of cue–target predictability on cueing effects. A group of participants received instructions about the informative value of the cue, while another group did not receive such instructions. Participants were able to extract the predictive value of a spatially peripheral cue and use it to orient attention, whether or not specific instructions about the predictive value of the cue were given, and no matter their ability to correctly report it in a post-test questionnaire. In the non-predictive block, bad estimators who received no instructions showed regular cueing effects, while good estimators exhibited smaller and non-significant facilitatory effects at the short SOA and an absence of significant IOR at longer SOAs. However, for the instructions group, the pattern of results reversed.     

Un regard structural sur l’inégalité des genres au travail : La causalité des différences « genrées » dans les caractéristiques de réseau de l’employé

This study aims at identifying gender differences in network structure and, hence, at understanding the underlying nature of gender inequalities in organisations. The hypotheses address four ego-network characteristics: network size and density, and ego’s centrality and brokerage. Contrary to the existing body of literature, none significant isolated gender difference in network structure is identified. Instead, those gender differences are triggered exclusively by organisational status, that is to say, they are due to the increased level of power. Thus, organisational status is considered as moderating the relationship between gender and network structure’s characteristics. Some further perspectives are suggested to clarify and generalise the phenomena observed.     

The Isolation, Primacy, and Recency Effects Predicted by an Adaptive LTD/LTP Threshold in Postsynaptic Cells

An item that stands out (is isolated) from its context is better remembered than an item consistent with the context. This isolation effect cannot be accounted for by increased attention, because it occurs when the isolated item is presented as the first item, or by impoverished memory of nonisolated items, because the isolated item is better remembered than a control list consisting of equally different items. The isolation effect is seldom experimentally or theoretically related to the primacy or the recency effects—that is, the improved performance on the first few and last items, respectively, on the serial position curve. The primacy effect cannot easily be accounted for by rehearsal in short-term memory because it occurs when rehearsal is eliminated. This article suggests that the primacy, the recency, and the isolation effects can be accounted for by experience-dependent synaptic plasticity in neural cells. Neurological empirical data suggest that the threshold that determines whether cells will show long-term potentiation (LTP) or long-term depression (LTD) varies as a function of recent postsynaptic activity and that synaptic plasticity is bounded. By implementing an adaptive LTP-LTD threshold in an artificial neural network, the various aspects of the isolation, the primacy, and the recency effects are accounted for, whereas none of these phenomena are accounted for if the threshold is constant. This theory suggests a possible link between the cognitive and the neurological levels.     

Examining the Role of Social Network Intervention as an Integral Component of Community-Based,Family-Focused Practice

Social network intervention aimed at bolstering the informal supports of high risk families is recognized as a common element of community-based, family-focused practice models, such as intensive family preservation services (IFPS), multisystemic therapy (MST), and the wraparound process. The empirical research basis for these practice models is examined, with an eye toward discerning the extent to which network intervention is identified as a critical component of comprehensive service. Results reveal that few studies make clear the degree to which treatment adhered to a focus on natural network enhancement or the development of informal resources. Those that did, suggest that relatively few families received this form of ecologically-oriented intervention. Consequently, this body of practice research offers insufficient evidence of the benefits or limitations associated with network facilitation with multi-need families. Explanations for these findings are explored and directions for future research are recommended.     

Disclosure of HIV-Positive Status to Latino Gay Men's Social Networks

This study explored disclosure of serostatus in a sample of 155 HIV-positive Latino gay men from New York City and Washington, DC. We examined rates of disclosure to different members of the social network: mothers, fathers, close friends, and primary sexual partners. There were high rates of disclosure of HIV-positive serostatus to main partners and closest friends and lower rates to fathers and mothers. We examined the role of 3 contextual target-dependent factors (emotional closeness to target, anticipated reactions from target, and target's knowledge of sexual orientation), as well as acculturation and time since diagnosis. Three separate logistic regression models were performed to predict disclosure of HIV-positive status to 3 targets: mothers, fathers, and closest friends. We found that disclosure was not a generalized tendency, but rather different factors were influential depending on the target. Whether the target was aware of participant's sexual orientation was associated with disclosure in all 3 models. Greater emotional closeness also predicted disclosure to mother and father; greater U.S. acculturation was associated with disclosure to father and marginally to mother. A longer time since diagnosis was associated with disclosure to the closest friend. These findings highlight the importance of taking into account roles and relationships, and their effect on disclosure.     

Histrionic Personality Disorder, Physical Attractiveness, and Social Adjustment

Histrionic personality disorder (HPD) is the only DSM-IV personality disorder (PD) explicitly linked to a person's physical appearance. This study examined the HPD–attractiveness link in a mixed-sex sample of college students. Consistent with expectations, HPD women were rated higher in attractiveness than women with other PDs or no PD. However, a parallel HPD–attractiveness link was not found in men. Subsequent analyses indicated that, relative to less attractive HPD women, more attractive HPD women (a) had a more varied and supportive social network, (b) exhibited more negative behaviors in important relationships, and (c) showed greater use of immature defenses, and less reliance on image-distorting, self-sacrificing, and mature defenses. Similar attractiveness–defense relationships were found in HPD men, but parallel results did not emerge for men in the other two domains. Implications of these findings are discussed, and suggestions for future studies are offered.