Classification and Categorical Inputs with Treed Gaussian Process Models

Recognizing the successes of treed Gaussian process (TGP) models as an interpretable and thrifty model for nonparametric regression, we seek to extend the model to classification. Both treed models and Gaussian processes (GPs) have, separately, enjoyed great success in application to classification problems. An example of the former is Bayesian CART. In the latter, real-valued GP output may be utilized for classification via latent variables, which provide classification rules by means of a softmax function. We formulate a Bayesian model averaging scheme to combine these two models and describe a Monte Carlo method for sampling from the full posterior distribution with joint proposals for the tree topology and the GP parameters corresponding to latent variables at the leaves. We concentrate on efficient sampling of the latent variables, which is important to obtain good mixing in the expanded parameter space. The tree structure is particularly helpful for this task and also for developing an efficient scheme for handling categorical predictors, which commonly arise in classification problems. Our proposed classification TGP (CTGP) methodology is illustrated on a collection of synthetic and real data sets. We assess performance relative to existing methods and thereby show how CTGP is highly flexible, offers tractable inference, produces rules that are easy to interpret, and performs well out of sample.     

MCS: A Method for Finding the Number of Clusters

This paper proposes a maximum clustering similarity (MCS) method for determining the number of clusters in a data set by studying the behavior of similarity indices comparing two (of several) clustering methods. The similarity between the two clusterings is calculated at the same number of clusters, using the indices of Rand (R), Fowlkes and Mallows (FM), and Kulczynski (K) each corrected for chance agreement. The number of clusters at which the index attains its maximum is a candidate for the optimal number of clusters. The proposed method is applied to simulated bivariate normal data, and further extended for use in circular data. Its performance is compared to the criteria discussed in Tibshirani, Walther, and Hastie (2001). The proposed method is not based on any distributional or data assumption which makes it widely applicable to any type of data that can be clustered using at least two clustering algorithms.     

On Definitions of Information in Physics

During the refereeing procedure of Anthropomorphic Quantum Darwinism by Thomas Durt, it became apparent in the dialogue between him and me that the definition of information in Physics is something about which not all authors agreed. This text aims at describing the concepts associated to information that are accepted as the standard in the Physics world community.     

Displacing Epistemology: Being in the Midst of Technoscientific Practice

Interest the Erklären?CVerstehen debate is usually interpreted as primarily epistemological. By raising the possibility that there are fundamentally different methods for fundamentally different types of science, the debate puts into play all the standard issues??that is, issues concerning scientific explanation and justification, the unity and diversity of scientific disciplines, the reality of their subject matter, the accessibility of various subject matters to research, and so on. In this paper, however, I do not focus on any of these specific issues. I start instead from the fact that the very existence of the debate itself is an issue; in fact, it poses a philosophical problem that almost everyone but the hardest line logical empiricists has come to realize cannot be resolved epistemologically. In my view, however, that it cannot be resolved ontologically, either. I think the problem is at bottom hermeneutical, and its resolution requires that we focus first, not on the objects of science or the methods of studying them, but on the character of the philosophical orientation assumed by those who would try to resolve it. In this paper, I explain why I think this is so by analyzing (1) Dilthey??s contribution to the original debate, (2) Husserl??s reaction to Dilthey, and (3) Heidegger??s critical evaluation of both. This line of philosophical development??this movement of self-understanding from critiques of objectivism to hermeneutical phenomenology??is of course already a central feature of much work in continental philosophy of science. In my conclusion, however, I argue for the less well-established??even if apparently approved??idea that it ought to be a central feature of technoscience studies as well.     

Observed increase in local cooling effect of deforestation at higher latitudes

Deforestation in mid- to high latitudes is hypothesized to have the potential to cool the Earth's surface by altering biophysical processes. In climate models of continental-scale land clearing, the cooling is triggered by increases in surface albedo and is reinforced by a land albedo-sea ice feedback. This feedback is crucial in the model predictions; without it other biophysical processes may overwhelm the albedo effect to generate warming instead. Ongoing land-use activities, such as land management for climate mitigation, are occurring at local scales (hectares) presumably too small to generate the feedback, and it is not known whether the intrinsic biophysical mechanism on its own can change the surface temperature in a consistent manner. Nor has the effect of deforestation on climate been demonstrated over large areas from direct observations. Here we show that surface air temperature is lower in open land than in nearby forested land. The effect is 0.85 ± 0.44 K (mean ± one standard deviation) northwards of 45° N and 0.21 ± 0.53 K southwards. Below 35° N there is weak evidence that deforestation leads to warming. Results are based on comparisons of temperature at forested eddy covariance towers in the USA and Canada and, as a proxy for small areas of cleared land, nearby surface weather stations. Night-time temperature changes unrelated to changes in surface albedo are an important contributor to the overall cooling effect. The observed latitudinal dependence is consistent with theoretical expectation of changes in energy loss from convection and radiation across latitudes in both the daytime and night-time phase of the diurnal cycle, the latter of which remains uncertain in climate models.     

Comprehensive analysis of the chromatin landscape in Drosophila melanogaster

Chromatin is composed of DNA and a variety of modified histones and non-histone proteins, which have an impact on cell differentiation, gene regulation and other key cellular processes. Here we present a genome-wide chromatin landscape for Drosophila melanogaster based on eighteen histone modifications, summarized by nine prevalent combinatorial patterns. Integrative analysis with other data (non-histone chromatin proteins, DNase I hypersensitivity, GRO-Seq reads produced by engaged polymerase, short/long RNA products) reveals discrete characteristics of chromosomes, genes, regulatory elements and other functional domains. We find that active genes display distinct chromatin signatures that are correlated with disparate gene lengths, exon patterns, regulatory functions and genomic contexts. We also demonstrate a diversity of signatures among Polycomb targets that include a subset with paused polymerase. This systematic profiling and integrative analysis of chromatin signatures provides insights into how genomic elements are regulated, and will serve as a resource for future experimental investigations of genome structure and function.     

Microwave quantum logic gates for trapped ions

Control over physical systems at the quantum level is important in fields as diverse as metrology, information processing, simulation and chemistry. For trapped atomic ions, the quantized motional and internal degrees of freedom can be coherently manipulated with laser light. Similar control is difficult to achieve with radio-frequency or microwave radiation: the essential coupling between internal degrees of freedom and motion requires significant field changes over the extent of the atoms' motion, but such changes are negligible at these frequencies for freely propagating fields. An exception is in the near field of microwave currents in structures smaller than the free-space wavelength, where stronger gradients can be generated. Here we first manipulate coherently (on timescales of 20 nanoseconds) the internal quantum states of ions held in a microfabricated trap. The controlling magnetic fields are generated by microwave currents in electrodes that are integrated into the trap structure. We also generate entanglement between the internal degrees of freedom of two atoms with a gate operation suitable for general quantum computation; the entangled state has a fidelity of 0.76(3), where the uncertainty denotes standard error of the mean. Our approach, which involves integrating the quantum control mechanism into the trapping device in a scalable manner, could be applied to quantum information processing, simulation and spectroscopy.     

Oxysterols direct immune cell migration via EBI2

Epstein-Barr virus-induced gene 2 (EBI2, also known as GPR183) is a G-protein-coupled receptor that is required for humoral immune responses; polymorphisms in the receptor have been associated with inflammatory autoimmune diseases. The natural ligand for EBI2 has been unknown. Here we describe the identification of 7α,25-dihydroxycholesterol (also called 7α,25-OHC or 5-cholesten-3β,7α,25-triol) as a potent and selective agonist of EBI2. Functional activation of human EBI2 by 7α,25-OHC and closely related oxysterols was verified by monitoring second messenger readouts and saturable, high-affinity radioligand binding. Furthermore, we find that 7α,25-OHC and closely related oxysterols act as chemoattractants for immune cells expressing EBI2 by directing cell migration in vitro and in vivo. A critical enzyme required for the generation of 7α,25-OHC is cholesterol 25-hydroxylase (CH25H). Similar to EBI2 receptor knockout mice, mice deficient in CH25H fail to position activated B cells within the spleen to the outer follicle and mount a reduced plasma cell response after an immune challenge. This demonstrates that CH25H generates EBI2 biological activity in vivo and indicates that the EBI2-oxysterol signalling pathway has an important role in the adaptive immune response.