Action Requests Research Articles

Optimal decision making with biologically realistic neural signals

Action selection in animals requires rapid decision making that can discriminate the most salient requests for behavioral expression. The basal ganglia (BG) are believed to play a critical role in resolving competition between these requests [1] and, recently it has been proposed that the BG and cortex, taken together, implement a decision algorithm known as the multi-hypothesis sequential probability ratio test (MSPRT) [2]. Here, the cortex first integrates noisy ‘evidence’ indicating salience of action requests. The BG then examine this integrated evidence, and report the channel with maximal mean salience. The MSPRT is optimal in the sense that it guarantees the smallest decision time for a given error. The signals used in [2] assume cortical evidence - as neural firing rates - is supplied by Gaussian distributed signals. However, it is known that such rates are not distributed normally. To remedy this we have deduced three new forms for MSPRT that are based on more biologically realistic input signals. They include the Inverse Gaussian, Gamma and Lognormal distributions. All implementations preserve a form which could be mapped, in principle, onto the anatomy of the BG. Figure ​Figure11 shows an example of the performance of the MSPRT models across all distributions, together with the so called race model. All variants used the same input statistics (mean and variance). Counterintuitively, the relative performance of the models did not change with base level of salience or evidence. Figure 1 Performance of original Gaussian MSPRT[2], new implementations of MSPRT, and Race model. Performance is measured as decision time against number of alternatives for selection. The selection criterion was 1% error rate. The results are averaged across ... A critical parameter in the algorithm is the time step Δt used to sample the cortical input. The Gaussian, Inverse Gaussian, and Gamma cases are generated as stable Levy processes and may be scaled indefinitely with decreasing Δt. However, unlike the MSPRT with Gaussian signals, we show that decision time for the inverse Gaussian, and Gamma decreases with Δt. Figure ​Figure11 use a Δt = 1 ms for all models and clearly shows a performance advantage for the cases with more realistic distributions (the lognormal case is shown with similar scaling even though it has not an associated Levy process). Indefinitely small decision times may appear unrealistic, but we present an interpretation of cortical sampling that relates Δt to the inter-spike intervals of an ensemble of neural afferents impinging on the cortical ‘integrators’. This and, other mechanisms, provide a natural lower bound for Δt. We conclude that neurons may take advantage of the properties of natural spike trains to enhance decision making in the cortex-BG complex.

Coordination Protocols for a Reliable Sensor, Actuator, and Device Network (SADN)

A sensor, actuator, and device network (SADN) is composed of three types of nodes, which are sensor, actuator, and actuation device nodes. Sensor nodes and actuator nodes are interconnected in wireless networks as discussed in wireless sensor and actuator networks (WSANs). Actuator nodes and device nodes are interconnected in types of networks, i.e. wireless and wired network. Sensor nodes sense an physical event and send sensed values of the event to actuator nodes. An actuator node makes a decision on proper actions on receipt of sensed values and then issue the action requests to the device nodes. A device node really acts to the physical world. For example, moves a robot arms by performing the action on receipt of the action request. Messages may be lost and nodes may be faulty. Especially, messages are lost due to noise and collision in a wireless network. We propose a fully redundant model for an SADN where each of sensor, actuator, and device functions is replicated in multiple nodes and each of sensor‐actuator and actuator‐device communication is realized in many‐to‐many type of communication protocols. Even if some number of nodes are faulty, the other nodes can perform requested tasks. Here, each sensor node sends sensed values to multiple actuator nodes and each actuator node receives sensed values from multiple sensor nodes. While multiple actuator nodes communicate with multiple replica nodes of a device. Even if messages are lost and some number of nodes are faulty, device nodes can surely receive action requests required for sensed values and the actions are performed. In this paper, we discuss a type of semi‐passive coordination (SPC) protocol of multiple actuator nodes for multiple sensor nodes. We discuss a type of active coordination protocol for multiple actuator nodes and multiple actuation device nodes. We evaluate the SPC protocol for the sensor‐actuator coordination in terms of the number of messages exchanged among actuators.

Speech acts, social context and parent-toddler play between the ages of 1;5 and 2;3

This study attempts to show that pragmatic development in young children—in terms of speech acts—involves the gradual acquisition of the capacity to take several dimensions of the interaction situation into account simultaneously. Communicative behavior patterns (frequency of different types of speech acts) are studied in relation to two components of the interaction context: the characteristics of the child's conversational partner (father or mother) and the type of play. Three groups of ten French urban middle class toddlers whose mean ages were 1;5, 1;10, and 2;3 were observed as they interacted with each parent. Materials for object building or free play were provided. The results showed that by the age of 1;5, the parent's gender had a significant effect on the production of directives, and then later, on that of assertives. The type of play started to have an impact at 22 months, affecting action requests and expressives (more frequent in object building), as well as requests for information (more frequent in free play). At the age of 2;3, these two factors had a joint effect on the speech act distribution: action requests were the most prevalent when the child was building a pre-specified object with the father. These developmental processes are discussed in terms of interaction “formats”, as described by J. Bruner.

What do facial expressions convey: feeling states, behavioral intentions, or action requests?

Emotion theorists assume certain facial displays to convey information about the expresser's emotional state. In contrast, behavioral ecologists assume them to indicate behavioral intentions or action requests. To test these contrasting positions, over 2,000 online participants were presented with facial expressions and asked what they revealed-feeling states, behavioral intentions, or action requests. The majority of the observers chose feeling states as the message of facial expressions of disgust, fear, sadness, happiness, and surprise, supporting the emotions view. Only the anger display tended to elicit more choices of behavioral intention or action request, partially supporting the behavioral ecology view. The results support the view that facial expressions communicate emotions, with emotions being multicomponential phenomena that comprise feelings, intentions, and wishes.

Effect of request type and situational features on negative politeness in requests

Politeness Theory (Brown & Levinson, 1987) implies that the negative politeness in requests should increase with the degree of threat to the hearer's negative face. The present study asked whether this effect depends on the type of request being made. Participants produced either borrow requests or stop action requests in situations that involved either a high threat or low threat to the hearer's face. As hypothesized, greater face threat produced more negative politeness (apologies and promises) in borrow requests but not in stop action requests.

Young children's sensitivity to direct and indirect request structure

Effects of request structure and listener characteristics on young children's compliance with requests were examined in two experi ments. In Experiment 1, an adult used direct (imperative) requests (e.g., 'Give me ...'/'What is ... ?'), indirect embedded requests (e.g., 'Can you give me ... ?'/`Can you tell me ... ?'), and hint requests (e.g., 'The water is spilling ...'/`I wonder what ...') for action and information. The requests were used with 3- and 5-year- olds (10 at each age) in structured play interactions. Five-year-olds complied with all request forms more often than 3-year-olds. Within action requests, both groups complied with the embedded form most often, followed by the direct then hint forms. Within information requests, they complied with the direct form most often, followed by embedded then hint forms. In Experiment 2, syntactically simple and complex direct and embedded forms were used with 4-year-olds grouped for tested language comprehension ability. The 15 low-ability children complied most frequently with simple direct and simple action requests. The 15 high-ability children surpassed low-ability children in compliance with complex direct, simple embedded, complex action, and simple and complex information requests. In contrast with previous studies, our findings show differential responsiveness to requests as a function of request structure and syntactic complexity as well as age.

BRE: A conversational program for bivariate regression and eta

Description. The regression program described here is unique in that it is a small self-contained package that also calculates eta (the ratio). one-way analysis of variance, a graph ofthe group means. and provides for limited data manipulation capability such as recoding. The organization and structure of the program is oriented totally to conversational interaction; it is designed for use in timesharing or on-line environments. The mix of statistical and data tasks that are available within the program suggest three major uses: (I) detection of nonlinear relationships by comparison of eta with the linear regression of x on y, (2) exploration of the impact of coding/recoding of variable x upon the calculation of eta and the one-way analysis of variance, and (3) computation of a monotonic variant of the ratio (see Leik & Gove, 1969). Leik and Gove show how Eta can be interpreted as an index of monotonicity. but only when the group means are monotonic and the group variances are homoscedastic. The recoding feature of the ORE program provides the potential for rearranging the x variable coding so as to achieve monotonicity. The following commands or action requests are available: (I) HELP: Instructions for conversational use are printed. (2) ETA: Both the bivariate regression and eta are computed for any pair of xy variables. Eta. also called the correlation ratio. is calculated as specified in common statistical texts (e.g .• Anderson & Zelditch, 1968; Edwards. 1970; Hays. 1973; Kerlinger, 1973; Loether & MeTavish, 1974; McNemar. 1962; & Mueller. Scheussler, & Costner. 1970). The algorithm involves sorting on x before the mean of y is computed for each group on x, The mean and other statistics are printed for each group so the printout is extensive when there are many categories or groups on x. A graphical display of the group means and variances is optional. (3) LIST: Any number of cases of the data matrix are printed. This is especially useful to check on the effects of recoding. (4) RECODE: Values on any variable are