Estimating Counterfactual Treatment Outcomes Over Time in Complex Multiagent Scenarios.

Evaluation of intervention in a multiagent system, for example, when humans should intervene in autonomous driving systems and when a player should pass to teammates for a good shot, is challenging in various engineering and scientific fields. Estimating the individual treatment effect (ITE) using counterfactual long-term prediction is practical to evaluate such interventions. However, most of the conventional frameworks did not consider the time-varying complex structure of multiagent relationships and covariate counterfactual prediction. This may lead to erroneous assessments of ITE and difficulty in interpretation. Here, we propose an interpretable, counterfactual recurrent network in multiagent systems to estimate the effect of the intervention. Our model leverages graph variational recurrent neural networks (GVRNNs) and theory-based computation with domain knowledge for the ITE estimation framework based on long-term prediction of multiagent covariates and outcomes, which can confirm the circumstances under which the intervention is effective. On simulated models of an automated vehicle and biological agents with time-varying confounders, we show that our methods achieved lower estimation errors in counterfactual covariates and the most effective treatment timing than the baselines. Furthermore, using real basketball data, our methods performed realistic counterfactual predictions and evaluated the counterfactual passes in shot scenarios.

Synthesis, structure, and magnetic properties of dinuclear cobalt(II) complexes with a new phenol-based dinucleating ligand with four hydroxyethyl chelating arms.

A new end-off type acyclic ligand with four hydroxyethyl arms, 2,6-bis[bis(2-hydroxyethyl)aminomethyl]-4-methylphenol [H(bhmp)], formed dinuclear cobalt(II) complexes [Co(2)(bhmp)(OAc)(2)]BPh(4) (1) and [Co(2)(bhmp)(OBz)(2)]BPh(4) (2). The complex 1.2.5CH(3)CN (C(50)H(62.5)BCo(2)N(4.5)O(9)) crystallizes in the monoclinic space group C2/c with dimensions a = 25.424(5) A, b = 13.376(2) A, c = 29.913(6) A, beta = 105.930(3) degrees, and V = 9781(3) A(3) and with Z = 8. X-ray diffraction analysis revealed a mu-phenoxo-bis(mu-acetato)dicobalt(II) core structure containing two octahedral cobalt(II) ions. Electronic spectra were investigated for 1 and 2 in the range 400-1800 nm, and the data were typical for the octahedral high-spin cobalt(II) complexes. Magnetic susceptibility was measured for 1 and 2 over the temperature range 4.5-300 K, and the data were analyzed well using our theoretical method. The best fitting parameters were kappa = 0.77, lambda = -116 cm(-1), Delta = 572 cm(-1), and J = -0.44 cm(-1) for complex 1 and kappa = 0.96, lambda = -93 cm(-1), Delta = 616 cm(-1), and J = -0.33 cm(-1) for complex 2.