Differentiating the Multipoint Expected Improvement for Optimal Batch Design

This work deals with parallel optimization of expensive objective functions which are modelled as sample realizations of Gaussian processes. The study is formalized as a Bayesian optimization problem, or continuous multi-armed bandit problem, where a batch of q > 0 arms is pulled in parallel at each iteration. Several algorithms have been developed for choosing batches by trading off exploitation and exploration. As of today, the maximum Expected Improvement (EI) and Upper Confidence Bound (UCB) selection rules appear as the most prominent approaches for batch selection. Here, we build upon recent work on the multipoint Expected Improvement criterion, for which an analytic expansion relying on Tallis’ formula was recently established. The computational burden of this selection rule being still an issue in application, we derive a closed-form expression for the gradient of the multipoint Expected Improvement, which aims at facilitating its maximization using gradient-based ascent algorithms. Substantial computational savings are shown in application. In addition, our algorithms are tested numerically and compared to state-of-the-art UCB-based batchsequential algorithms. Combining starting designs relying on UCB with gradient-based EI local optimization finally appears as a sound option for batch design in distributed Gaussian Process optimization.

Nonparametric location estimators in the randomized complete block design

Several tests for the comparison of different groups in the randomized complete block design exist. However, there is a lack of robust estimators for the location difference between one group and all the others on the original scale. The relative marginal effects are commonly used in this situation, but they are more difficult to interpret and use by less experienced people because of the different scale. In this paper two nonparametric estimators for the comparison of one group against the others in the randomized complete block design will be presented. Theoretical results such as asymptotic normality, consistency, translation invariance, scale preservation, unbiasedness, and median unbiasedness are derived. The finite sample behavior of these estimators is derived by simulations of different scenarios. In addition, possible confidence intervals with these estimators are discussed and their behavior derived also by simulations.

Results of the Questionnaire

Early diagnosis of patients with compensated advanced chronic liver disease (cACLD) and portal hypertension is challenging in clinical practice. A growing amount of evidence regarding noninvasive diagnostic methods, and in particular liver stiffness measurement (LSM), suggests that these tools could be used in clinical practice and might potentially limit the use of invasive, reference diagnostic tools (HVPG measurement and endoscopy). Our panel aimed at better understanding the opinion of the Baveno faculty regarding the current practice and use of invasive and noninvasive methods in the field of screening and surveillance of varices; a specific questionnaire was electronically sent to all the faculty members. The results suggested that the experts agreed on the use of noninvasive methods to rule out/identify patients with cACLD. They also indicated that the persistence or removal of the causal agent which led to cirrhosis should guide the choice of using the shortest or the longest interval among those recommended for surveillance endoscopies. Finally, the use of noninvasive methods in these clinical scenarios was pointed out as a relevant field for future research.

Consensus Statements: Session 1 - Screening and Surveillance

The introduction of transient elastography in clinical practice has allowed the early identification of patients with chronic liver disease (CLD) at risk of developing clinically significant portal hypertension (CSPH) (1b;A).

The Complexity of Non-Iterated Probabilistic Justification Logic

The logic PJ is a probabilistic logic defined by adding (noniterated) probability operators to the basic justification logic J. In this paper we establish upper and lower bounds for the complexity of the derivability problem in the logic PJ. The main result of the paper is that the complexity of the derivability problem in PJ remains the same as the complexity of the derivability problem in the underlying logic J, which is π[p/2] -complete. This implies that the probability operators do not increase the complexity of the logic, although they arguably enrich the expressiveness of the language.

Probabilistic Justification Logic

We present a probabilistic justification logic, PPJ, to study rational belief, degrees of belief and justifications. We establish soundness and completeness for PPJ and show that its satisfiability problem is decidable. In the last part we use PPJ to provide a solution to the lottery paradox.

Leaving Home: Philo of Alexandria on the Exodus.

For Jewish-Hellenistic authors writing in Egypt, the Exodus story posed unique challenges. After all, to them Egypt was, as Philo of Alexandria states, their fatherland. How do these authors come to terms with the biblical story of liberation from Egyptian slavery and the longing for the promised land? In this chapter I am taking a close look at Philo’s detailed discussion of the Exodus and locate it within the larger context of Jewish-Hellenistic literature (Wisdom of Solomon, Ezekiel’s Exagoge). In Philo’s rewriting of the Exodus the destination of the journey is barely mentioned. Contrary to the biblical narrative, in the scene of the burning bush, as retold by Philo, God does not tell Moses where to go. Philo’s main concern is what happens in Egypt: both in biblical times and in his own days. The Exodus is nevertheless important to Philo: He reads the story allegorically as a journey from the land of the body to the realms of the mind. Such a symbolic reading permitted him to control the meaning of the Exodus and to stay, literally and figuratively, in Egypt.

Biocommunication of Ciliates

This is the first coherent description of all levels of communication of ciliates. Ciliates are highly sensitive organisms that actively compete for environmental resources. They assess their surroundings, estimate how much energy they need for particular goals, and then realise the optimum variant. They take measures to control certain environmental resources. They perceive themselves and can distinguish between ‘self’ and ‘non-self’. They process and evaluate information and then modify their behaviour accordingly. These highly diverse competences show us that this is possible owing to sign(aling)-mediated communication processes within ciliates (intra-organismic), between the same, related and different ciliate species (inter-organismic), and between ciliates and non-ciliate organisms (trans-organismic). This is crucial in coordinating growth and development, shape and dynamics. This book further serves as a learning tool for research aspects in biocommunication in ciliates. It will guide scientists in further investigations on ciliate behavior, how they mediate signaling processes between themselves and the environment.

RNA-Guided Genome Editing

Ciliates have evolved highly complex and intricately controlled pathways to ensure the precise and complete removal of all genomic sequences not required for vegetative growth. At the same time, they retain a reference copy of all their genetic information for future generations. This chapter describes how different ciliates use RNA-mediated DNA comparison processes to form new somatic nuclei from germline nuclei. While these processes vary in their precise mechanisms, they all use RNA to target genomic DNA sequences—either for retention or elimination. They also all consist of more than one individual pathway acting cooperatively—the two subsets of small RNAs in Paramecium and the guide RNAs and Piwi-interacting RNAs in Oxytricha—to ensure a strong belt-and-braces approach to consistent and precise somatic nucleus development. Nonetheless, this genome comparison approach to somatic nucleus development provides an elegant method for trans-generational environmental adaptation. Conceptually, it is easy to imagine how somatic changes that occur during vegetative growth could be transferred to meiotic offspring, while an unaltered germline genome is retained. Further research in this area will have far-reaching implications for the trans-generational adaptation of more distantly related eukaryotes, such as humans.

Eine Kundgebung des Chacham-Baschi von Tiberias

Cham-Baschi Abulafia