Marked point processes and intensity ratios for limit order book modeling

This paper extends the analysis of Muni Toke and Yoshida (2020) to the case of marked point processes. We consider multiple marked point processes with intensities defined by three multiplicative components, namely a common baseline intensity, a state-dependent component specific to each process, and a state-dependent component specific to each mark within each process. We show that for specific mark distributions, this model is a combination of the ratio models defined in Muni Toke and Yoshida (2020). We prove convergence results for the quasi-maximum and quasi-Bayesian likelihood estimators of this model and provide numerical illustrations of the asymptotic variances. We use these ratio processes to model transactions occurring in a limit order book. Model flexibility allows us to investigate both state-dependency (emphasizing the role of imbalance and spread as significant signals) and clustering. Calibration, model selection and prediction results are reported for high-frequency trading data on multiple stocks traded on Euronext Paris. We show that the marked ratio model outperforms other intensity-based methods (such as “pure” Hawkes-based methods) in predicting the sign and aggressiveness of market orders on financial markets.

Mechanisms of Memory Updating: State Dependency vs. Reconsolidation

Reactivating a memory trace has been argued to put it in a fragile state where it must undergo a stabilization process known as reconsolidation. During this process, memories are thought to be susceptible to interference and can be updated with new information. In the spatial context paradigm, memory updating has been shown to occur when new information is presented in the same spatial context as old information, an effect attributed to a reconsolidation process. However, the integration concept holds that memory change can only occur when reactivation and test states are the same, similar to a state-dependent effect. Thus, in human episodic memory, memory updating should only be found when state is the same across the study, reactivation, and test sessions. We investigated whether memory updating can be attributed to state dependency in two experiments using mood as a state. We found evidence of memory updating only when mood was the same across all sessions of the experiments, lending support to the integration concept and posing a challenge to a reconsolidation explanation.

Local brain-state dependency of effective connectivity: evidence from TMS–EEG

Background: Spontaneous cortical oscillations have been shown to modulate cortical responses to transcranial magnetic stimulation (TMS). If not controlled for, they might increase variability in responses and mask meaningful changes in the signals of interest when studying the brain with TMS combined with electroencephalography (TMS–EEG). To address this challenge in future closed-loop stimulation paradigms, we need to understand how spontaneous oscillations affect TMS-evoked responses. Objective: To describe the effect of the pre-stimulus phase of cortical mu (8–13 Hz) and beta (13–30 Hz) oscillations on TMS-induced effective connectivity patterns. Methods: We applied TMS to the left primary motor cortex and right pre-supplementary motor area of three subjects while recording EEG. We classified trials off-line into positive- and negative-phase classes according to the mu and beta rhythms. We calculated differences in the global mean-field amplitude (GMFA) and compared the cortical spreading of the TMS-evoked activity between the two classes. Results: Phase had significant effects on the GMFA in 11 out of 12 datasets (3 subjects × 2 stimulation sites × 2 frequency bands). Seven of the datasets showed significant differences in the time range 15–50 ms, nine in 50–150 ms, and eight after 150 ms post-stimulus. Source estimates showed complex spatial differences between the classes in the cortical spreading of the TMS-evoked activity. Conclusions: TMS-evoked effective connectivity appears to depend on the phase of local cortical oscillations at the stimulated site. This may be crucial for efficient design of future brain-state-dependent and closed-loop stimulation paradigms.

Isometric agonist and antagonist muscle activation interacts differently with 140 Hz tACS aftereffects at different intensities

Introduction: 1) During tES, increasing intracellular Ca2+ levels beyond those needed for inducing LTP may collapse aftereffects. 2) State-dependent plastic aftereffects are reduced when applied during muscle activation as compared to rest. 3) Cortical surround inhibition by antagonistic muscle activation inhibits the center-innervated agonist. Objectives: To determine the interaction of state dependency of tACS aftereffects at rest and under activation of agonist and antagonist muscles during stimulation with different intensities. Methods: In thirteen healthy participants, we measured MEP amplitudes before and after applying tACS at 140 Hz over the motor cortex in nine single-blinded sessions using sham, 1 mA and 2 mA stimulation intensities during rest and activation of agonist and antagonist muscles. Results: During rest, only 1 mA tACS produced a significant MEP increase, while the 2 mA stimulation produced no significant MEP size shift. During agonist activation 1 mA did not induce MEP changes, after 2 mA first a decrease and later an increase of MEPs were observed. Antagonist activation under sham tACS led to an inhibition, which was restored to baseline by 1 and 2 mA tACS. Conclusions: Increasing stimulation intensity beyond 1 mA does not increase excitability, compatible with too strong intracellular Ca2+increase. Antagonist innervation leads to MEP inhibition supporting the concept of surround inhibition, which can be overcome by tACS at both intensities. During agonist innervation a tACS dose dependent relationship exists. Significance: Our results integrate concepts of "leaky membranes" under activation, surround inhibition, intracellular Ca2+ increase and their role in the aftereffects of tACS.

State-dependent studies on perception and cognition

Neuronal response to an external stimulus is affected not only by stimulus properties, but also by the baseline activation state; this is referred to as state-dependency. Leveraging this principle helps to enhance the specificity and reduce the variability of brain stimulation effects. State-dependent paradigms have proven to be successful in enhancing the functional resolution of brain stimulation to the extent that the tuning of neuronal representations can be revealed, and they have also enhanced clinical benefits in the treatment of disorders such as depression. Furthermore, state-dependent approach has been applied in various brain stimulation protocols, including online and offline transcranial magnetic stimulation (TMS), transcranial direct current stimulation (tDCS), transcranial alternating current stimulation (tACS), transcranial random noise stimulation (tRNS), and paired-pulse associative stimulation. This chapter describes the principles and mechanisms of state-dependent brain stimulation and summarizes its contribution to cognitive neuroscience.