Force Myography-Based Human Robot Interactions via Deep Domain Adaptation and Generalization

Estimating applied force using force myography (FMG) technique can be effective in human-robot interactions (HRI) using data-driven models. A model predicts well when adequate training and evaluation are observed in same session, which is sometimes time consuming and impractical. In real scenarios, a pretrained transfer learning model predicting forces quickly once fine-tuned to target distribution would be a favorable choice and hence needs to be examined. Therefore, in this study a unified supervised FMG-based deep transfer learner (SFMG-DTL) model using CNN architecture was pretrained with multiple sessions FMG source data (Ds, Ts) and evaluated in estimating forces in separate target domains (Dt, Tt) via supervised domain adaptation (SDA) and supervised domain generalization (SDG). For SDA, case (i) intra-subject evaluation (Ds ≠ Dt-SDA, Ts ≈ Tt-SDA) was examined, while for SDG, case (ii) cross-subject evaluation (Ds ≠ Dt-SDG, Ts ≠ Tt-SDG) was examined. Fine tuning with few “target training data” calibrated the model effectively towards target adaptation. The proposed SFMG-DTL model performed better with higher estimation accuracies and lower errors (R2 ≥ 88%, NRMSE ≤ 0.6) in both cases. These results reveal that interactive force estimations via transfer learning will improve daily HRI experiences where “target training data” is limited, or faster adaptation is required.

Integrated Pharmacogenetics Analysis of the Three Fangjis Decoctions for Treating Arrhythmias Based on Molecular Network Patterns

Aim: To explore the diverse target distribution and variable mechanisms of different fangjis prescriptions when treating arrhythmias based on the systems pharmacology.Methods: The active ingredients and their corresponding targets were acquired from the three fangjis [Zhigancao Tang (ZT), Guizhigancao Longgumuli Tang (GLT), and Huanglian E'jiao Tang (HET)] and the arrhythmia-related genes were identified based on comprehensive database screening. Networks were constructed between the fangjis and arrhythmia and used to define arrhythmia modules. Common and differential gene targets were identified within the arrhythmia network modules and the cover rate (CR) matrix was applied to compare the contributions of the fangjis to the network and modules. Comparative pharmacogenetics analyses were then conducted to define the arrhythmia-related signaling pathways regulated by the fangjis prescriptions. Finally, the divergence and convergence points of the arrhythmia pathways were deciphered based on databases and the published literature.Results: A total of 187, 105, and 68 active ingredients and 1,139, 1,195, and 811 corresponding gene targets of the three fangjis were obtained and 102 arrhythmia-related genes were acquired. An arrhythmia network was constructed and subdivided into 4 modules. For the target distribution analysis, 65.4% of genes were regulated by the three fangjis within the arrhythmia network. ZT and GLT were more similar to each other, mainly regulated by module two, whereas HET was divided among all the modules. From the perspective of signal transduction, calcium-related pathways [calcium, cyclic guanosine 3′,5′-monophosphate (cGMP)-PKG, and cyclic adenosine 3′,5′-monophosphate (cAMP)] and endocrine system-related pathways (oxytocin signaling pathway and renin secretion pathways) were associated with all the three fangjis prescriptions. Nevertheless, heterogeneity existed between the biological processes and pathway distribution among the three prescriptions. GLT and HET were particularly inclined toward the conditions involving abnormal hormone secretion, whereas ZT tended toward renin-angiotensin-aldosterone system (RAAS) disorders. However, calcium signaling-related pathways prominently feature in the pharmacological activities of the decoctions. Experimental validation indicated that ZT, GLT, and HET significantly shortened the duration of ventricular arrhythmia (VA) and downregulated the expression of CALM2 and interleukin-6 (IL-6) messenger RNAs (mRNAs); GLT and HET downregulated the expression of CALM1 and NOS3 mRNAs; HET downregulated the expression of CRP mRNA.Conclusion: Comparing the various distributions of the three fangjis, pathways provide evidence with respect to precise applications toward individualized arrhythmia treatments.

Some New Tests of Conformity with Benford’s Law

This paper presents new perspectives and methodological instruments for verifying the validity of Benford’s law for a large given dataset. To this aim, we first propose new general tests for checking the statistical conformity of a given dataset with a generic target distribution; we also provide the explicit representation of the asymptotic distributions of the relevant test statistics. Then, we discuss the applicability of such novel devices to the case of Benford’s law. We implement extensive Monte Carlo simulations to investigate the size and the power of the introduced tests. Finally, we discuss the challenging theme of interpreting, in a statistically reliable way, the conformity between two distributions in the presence of a large number of observations.

High-dimensional distribution generation through deep neural networks

We show that every d-dimensional probability distribution of bounded support can be generated through deep ReLU networks out of a 1-dimensional uniform input distribution. What is more, this is possible without incurring a cost—in terms of approximation error measured in Wasserstein-distance—relative to generating the d-dimensional target distribution from d independent random variables. This is enabled by a vast generalization of the space-filling approach discovered in Bailey and Telgarsky (in: Bengio (eds) Advances in neural information processing systems vol 31, pp 6489–6499. Curran Associates, Inc., Red Hook, 2018). The construction we propose elicits the importance of network depth in driving the Wasserstein distance between the target distribution and its neural network approximation to zero. Finally, we find that, for histogram target distributions, the number of bits needed to encode the corresponding generative network equals the fundamental limit for encoding probability distributions as dictated by quantization theory.

Solutions of the Multivariate Inverse Frobenius–Perron Problem

We address the inverse Frobenius–Perron problem: given a prescribed target distribution ρ, find a deterministic map M such that iterations of M tend to ρ in distribution. We show that all solutions may be written in terms of a factorization that combines the forward and inverse Rosenblatt transformations with a uniform map; that is, a map under which the uniform distribution on the d-dimensional hypercube is invariant. Indeed, every solution is equivalent to the choice of a uniform map. We motivate this factorization via one-dimensional examples, and then use the factorization to present solutions in one and two dimensions induced by a range of uniform maps.

α-Geodesical Skew Divergence

The asymmetric skew divergence smooths one of the distributions by mixing it, to a degree determined by the parameter λ, with the other distribution. Such divergence is an approximation of the KL divergence that does not require the target distribution to be absolutely continuous with respect to the source distribution. In this paper, an information geometric generalization of the skew divergence called the α-geodesical skew divergence is proposed, and its properties are studied.

Generative Adversarial Network Performance in Low-Dimensional Settings

A generative adversarial network (GAN) is an artificial neural network with a distinctive training architecture, designed to createexamples that faithfully reproduce a target distribution. GANs have recently had particular success in applications involvinghigh-dimensional distributions in areas such as image processing. Little work has been reported for low dimensions, where properties of GANs may be better identified and understood. We studied GAN performance in simulated low-dimensional settings, allowing us totransparently assess effects of target distribution complexity and training data sample size on GAN performance in a simpleexperiment. This experiment revealed two important forms of GAN error, tail underfilling and bridge bias, where the latter is analogousto the tunneling observed in high-dimensional GANs.

$alpha$-Geodesical Skew Divergence

The asymmetric skew divergence smooths one of the distributions by mixing it, to a degree determined by the parameter $\lambda$, with the other distribution. Such divergence is an approximation of the KL divergence that does not require the target distribution to be absolutely continuous with respect to the source distribution. In this paper, an information geometric generalization of the skew divergence called the $\alpha$-geodesical skew divergence is proposed, and its properties are studied.

A private model for determining the efficiency of target distribution

A specific model has been developed to assess the performance indicators of target designation when setting obstacles by the enemy and maneuvering various targets. On the basis of the adopted technical characteristics and parameters of stations and weapon systems, optimal target designation options were found. Favorable conditions for firing rockets have also been determined. The indicator of the increase in the average number of destroyed targets is calculated when moving from one type of target placement to another. The course of solving a specific model is presented. Analytical expressions for calculating target indicators of placement efficiency and graphic material are given.