"Functional connectivity networks have become a central focus in neuroscience because they reveal key higher-dimensional features of normal and abnormal nervous system physiology. Functional networks reflect activity-based coupling between brain regions that may be constrained by relatively static anatomical connections, yet these networks appear to support tremendously dynamic behaviors. Within this growing field, the stability and temporal characteristics of functional connectivity brain networks have not been well characterized. We evaluated the temporal stability of spontaneous functional connectivity networks derived from multi-day scalp encephalogram (EEG) recordings in five healthy human subjects. Topological stability and graph characteristics of networks derived from averaged data epochs ranging from 1 s to multiple hours across different states of consciousness were compared. We show that, although functional networks are highly variable on the order of seconds, stable network templates emerge after as little as ∼100 s of recording and persist across different states and frequency bands (albeit with slightly different characteristics in different states and frequencies). Within these network templates, the most common edges are markedly consistent, constituting a network “core.” Although average network topologies persist across time, measures of global network connectivity, density and clustering coefficient, are state and frequency specific, with sparsest but most highly clustered networks seen during sleep and in the gamma frequency band. These findings support the notion that a core functional organization underlies spontaneous cortical processing and may provide a reference template on which unstable, transient, and rapidly adaptive long-range assemblies are overlaid in a frequency-dependent manner."

"The Neyman–Pearson lemma provides a simple procedure for optimally testing a single hypothesis when the null and alternative distributions are known. This result has played a major role in the development of significance testing strategies that are used in practice. Most of the work extending single-testing strategies to multiple tests has focused on formulating and estimating new types of significance measures, such as the false discovery rate. These methods tend to be based on p-values that are calculated from each test individually, ignoring information from the other tests. I show here that one can improve the overall performance of multiple significance tests by borrowing information across all the tests when assessing the relative significance of each one, rather than calculating p-values for each test individually. The ‘optimal discovery procedure’ is introduced, which shows how to maximize the number of expected true positive results for each fixed number of expected false positive results. The optimality that is achieved by this procedure is shown to be closely related to optimality in terms of the false discovery rate. The optimal discovery procedure motivates a new approach to testing multiple hypotheses, especially when the tests are related. As a simple example, a new simultaneous procedure for testing several normal means is defined; this is surprisingly demonstrated to outperform the optimal single-test procedure, showing that a method which is optimal for single tests may no longer be optimal for multiple tests. Connections to other concepts in statistics are discussed, including Stein's paradox, shrinkage estimation and the Bayesian approach to hypothesis testing."

"Various bootstraps have been proposed for bootstrapping clustered data from one-way arrays. The simulation results in the literature suggest that some of these methods work quite well in practice; the theoretical results are limited and more mixed in their conclusions. For example, McCullagh reached negative conclusions about the use of non-parametric bootstraps for one-way arrays. The purpose of this paper is to extend our understanding of the issues by discussing the effect of different ways of modelling clustered data, the criteria for successful bootstraps used in the literature and extending the theory from functions of the sample mean to include functions of the between and within sums of squares and non-parametric bootstraps to include model-based bootstraps. We determine that the consistency of variance estimates for a bootstrap method depends on the choice of model with the residual bootstrap giving consistency under the transformation model whereas the cluster bootstrap gives consistent estimates under both the transformation and the random-effect model. In addition we note that the criteria based on the distribution of the bootstrap observations are not really useful in assessing consistency."

"Why do some music styles gain mass popularity while others thrive in small niches? Banding Together explores this question and reveals the attributes that together explain the growth of twentieth-century American popular music. Drawing on a vast array of examples from sixty musical styles--ranging from rap and bluegrass to death metal and South Texas polka, and including several created outside the United States--Jennifer Lena uncovers the shared grammar that allows us to understand the cultural language and evolution of popular music.
"What are the common economic, organizational, ideological, and aesthetic traits among contemporary genres? Do genres follow patterns in their development? Lena discovers four dominant forms--Avant-garde, Scene-based, Industry-based, and Traditionalist--and two dominant trajectories that describe how American pop music genres develop. Outside the United States there exists a fifth form: the Government-purposed genre, which she examines in the music of China, Serbia, Nigeria, and Chile. Offering a rare analysis of how music communities operate, she looks at the shared obstacles and opportunities creative people face and reveals the ways in which people collaborate around ideas, artworks, individuals, and organizations that support their work."

Indeed. (Let me add that the idea of explaining the US current account deficit as imperial tribute is an idea which occurs to many bright, cynical 19 year olds, but just does not work when you think it through.)

"We present an online version of the expectation-maximization (EM) algorithm for hidden Markov models (HMMs). The sufficient statistics required for parameters estimation is computed recursively with time, that is, in an online way instead of using the batch forward-backward procedure. This computational scheme is generalized to the case where the model parameters can change with time by introducing a discount factor into the recurrence relations. The resulting algorithm is equivalent to the batch EM algorithm, for appropriate discount factor and scheduling of parameters update. On the other hand, the online algorithm is able to deal with dynamic environments, i.e., when the statistics of the observed data is changing with time. The implications of the online algorithm for probabilistic modeling in neuroscience are briefly discussed."

"We propose simple randomized strategies for sequential prediction under imperfect monitoring, that is, when the forecaster does not have access to the past outcomes but rather to a feedback signal. The proposed strategies are consistent in the sense that they achieve, asymptotically, the best possible average reward. It was Rustichini (1999) who first proved the existence of such consistent predictors. The forecasters presented here offer the first constructive proof of consistency. Moreover, the proposed algorithms are computationally efficient. We also establish upper bounds for the rates of convergence. In the case of deterministic feedback, these rates are optimal up to logarithmic terms."

"In this paper, we tackle the problem of prediction and confidence intervals for time series using a statistical learning approach and quantile loss functions. In a first time, we show that the Gibbs estimator (also known as Exponentially Weighted aggregate) is able to predict as well as the best predictor in a given family for a wide set of loss functions. In particular, using the quantile loss function of Koenker and Bassett (1978), this allows to build confidence intervals. We apply these results to the problem of prediction and confidence regions for the French Gross Domestic Product (GDP) growth, with promising results."

"In this paper we tackle the problem of fast rates in time series forecasting from a statistical learning perspective. In a serie of papers (e.g. Meir 2000, Modha and Masry 1998, Alquier and Wintenberger 2012) it is shown that the main tools used in learning theory with iid observations can be extended to the prediction of time series. The main message of these papers is that, given a family of predictors, we are able to build a new predictor that predicts the series as well as the best predictor in the family, up to a remainder of order $1/sqrt{n}$. It is known that this rate cannot be improved in general. In this paper, we show that in the particular case of the least square loss, and under a strong assumption on the time series (phi-mixing) the remainder is actually of order $1/n$. Thus, the optimal rate for iid variables, see e.g. Tsybakov 2003, and individual sequences, see cite{lugosi} is, for the first time, achieved for uniformly mixing processes. We also show that our method is optimal for aggregating sparse linear combinations of predictors."

"A defining feature of many large empirical networks is their intrinsic complexity. However, many networks also contain a large degree of structural repetition. An immediate question then arises: can we characterize essential network complexity while excluding structural redundancy? In this article we utilize inherent network symmetry to collapse all redundant information from a network, resulting in a coarse graining which we show to carry the essential structural information of the “parent” network. In the context of algebraic combinatorics, this coarse-graining is known as the “quotient.” We systematically explore the theoretical properties of network quotients and summarize key statistics of a variety of “real-world” quotients with respect to those of their parent networks. In particular, we find that quotients can be substantially smaller than their parent networks yet typically preserve various key functional properties such as complexity (heterogeneity and hub vertices) and communication (diameter and mean geodesic distance), suggesting that quotients constitute the essential structural skeletons of their parent networks. We summarize with a discussion of potential uses of quotients in analysis of biological regulatory networks and ways in which using quotients can reduce the computational complexity of network algorithms."

"A general self-consistency approach allows for a thorough treatment of the corrections to the mean-field approximation (MFA). The natural extension of standard MFA with the help of a cumulant expansion leads to a point of view on the effective field theories. The proposed approach can be used for a systematic treatment of fluctuation effects of various length scales and, perhaps, for the development of a coarse-graining procedure. We outline and justify our method by some preliminary calculations. Results are given for the critical temperature and the Landau parameters of the ϕ4 theory—the field counterpart of the Ising model. An important unresolved problem of the modern theory of phase transitions—the problem for the calculation of the true critical temperature—is considered within the framework of the present approach. A comprehensive description of the ground-state properties of many-body systems is also demonstrated."

"For independent $d$-variate random samples $X_1, cdots, X_{n_1}$ i.i.d. $f(x), Y_1, cdots, Y_{n_2}$ i.i.d. $g(x)$, where the densities $f$ and $g$ are assumed to be continuous a.e., consider the number $T$ of all $k$ nearest neighbor comparisons in which observations and their neighbors belong to the same sample. We show that, if $f = g$ a.e., the limiting (normal) distribution of $T$, as $min(n_1, n_2) rightarrow infty, n_1/(n_1 + n_2) rightarrow tau, 0 < tau < 1$, does not depend on $f$. An omnibus procedure for testing the hypothesis $H_0: f = g$ a.e. is obtained by rejecting $H_0$ for large values of $T$. The result applies to a general distance (generated by a norm on $mathbb{R}^d$) for determining nearest neighbors, and it generalizes to the multisample situation."

"Replication files":
http://www.chrisblattman.com/documents/data/shocks-conflict/Bazzi-Blattman.zip?9d7bd4

I want to see the trees!!!

Cf. http://www.theonion.com/content/node/57311

"distrust of traditional lending institutions such as commercial banks led some borrowers to seek out brokers from their own communities whom they felt they could trust. But these brokers were operating under high-powered incentives to inflate rates and fees and guide borrowers towards subprime products even when they were eligible for cheaper alternatives. The trust that was placed in the brokers allowed them greater flexibility to respond to these incentives and left borrowers worse off than the would have been if they had been more suspicious or better aware of the incentive structures in place.
"Viewed in this manner, the subprime saga has some broader implications. From the point of view of a company operating in multiple local markets with a diverse customer base, the strategy of giving local employees or contractors the discretion to adjust prices can be very profitable. This is especially so if these employees appear trustworthy to their customers, but are not in fact deserving of such trust. As Groucho Marx is reputed to have said: 'The secret of life is honesty and fair-dealing. If you can fake that you've got it made.' For products involving frequent repeat purchases by the same customer, reputation effects and competition can limit the degree of price discrimination. But the purchase of a home is an infrequent transaction for most people, and the complexity of the loan product precludes easy comparison with alternatives on offer. Trust then becomes a key determinant of pricing and transaction volume, especially when strong and hidden incentives for the betrayal of trust are in place.
"Betrayal also leads to the erosion of trust over time. It could be argued that trust is one of our most valuable public goods, substantially lowering the costs of transacting. In the complete absence of trust, the resources that would need to be devoted to monitoring would be prohibitive and many organizations and markets would simply not exist. Trust also comes naturally to most of us, based on simple cues such as those revealed in Reid's interviews. High-powered incentives to secure and then betray such trust are therefore costly not just to the immediate victim, but also to the community at large. This may be one of the less visible consequences of the subprime crisis."

"Aaaaaaand that’s then I want to start pressing the “It’s time for the goddamned revolution” button. By the time we get to the breakdowns of the monthly expenses of the seven 1% households profiled for the article, which features line items like $800 a month on wine and $1200 for the vacation house on the lake, I’m vaguely surprised Toronto isn’t on fire. The only people I feel any sort of commonality with are the immigrant family, who pack their own lunches for work and aside from the hair salon line item seem to have some perspective on their cash. The retired couple who invested well and are living off the proceeds also gets a pass, because, hey, that’s the goal, right? Otherwise: Purification by flame."

"I think he is right about these kids being comfortable with the “formal process” of applying to investment banks etc., but I don’t think he dives deep enough into why this is true. The fact is, the kids who get into Harvard nowadays are, generally speaking, professional test takers. They are moreover dependent on outside metrics for evaluating themselves. If you took away tests and grading systems, these kids would be desperately unhappy, because that’s how they’ve been trained all their lives to think about their self-worth.
"When I was a tutor at one of the undergrad houses at grad school, I was incredibly impressed with the international group of undergrads I was in charge of; their credentials, even at the age of 20, were amazing, and their knowledge and self-possession were stunning. Same with the high school kids I taught at math camp last summer. But one thing I saw time and time again was how much they needed to please some outside authority. It’s like they never decided whether they themselves liked their major or whether it was a good fit- it was instead about whether they’d be successful and whether it would be an impressive path for them. So, external metrics of success.
"Here’s my diagnosis. These kids are vulnerable to Wall Street investment firms and to things like Teach for America because they have application processes at all. But life, normal adult life, doesn’t have an application process. You actually, at some point, need to figure out what you want to do and what makes you happy. You need to take a leap of faith that your native talents and desires will end you up at a reasonable and interesting place.
"Actually you don’t ever have to decide that, you could just keep doing what you think looks good to other people and pleases your parents or friends, without regard to whether it fulfills you at all. That’s kind of what’s happening I think with the 36% of the Princeton undergrads going to finance."

"Math, as it happens, isn’t hard. Shopping is much more complicated."

"It is shown how to compute effective and functional connection matrices (eCMs and fCMs) from anatomical CMs (aCMs) and corresponding strength-of-connection matrices (sCMs) using propagator methods in which neural interactions play the role of scatterings. This analysis demonstrates how network effects dress the bare propagators (the sCMs) to yield effective propagators (the eCMs) that can be used to compute the covariances customarily used to define fCMs. The results incorporate excitatory and inhibitory connections, multiple structures and populations, asymmetries, time delays, and measurement effects. They can also be postprocessed in the same manner as experimental measurements for direct comparison with data and thereby give insights into the role of coarse-graining, thresholding, and other effects in determining the structure of CMs. The spatiotemporal results show how to generalize CMs to include time delays and how natural network modes give rise to long-range coherence at resonant frequencies. The results are demonstrated using tractable analytic cases via neural field theory of cortical and corticothalamic systems. These also demonstrate close connections between the structure of CMs and proximity to critical points of the system, highlight the importance of indirect links between brain regions and raise the possibility of imaging specific levels of indirect connectivity. Aside from the results presented explicitly here, the expression of the connections among aCMs, sCMs, eCMs, and fCMs in terms of propagators opens the way for propagator theory to be further applied to analysis of connectivity."

"Next month in Pittsburgh, Pennsylvania, artist Richard Pell opens the Center for PostNatural History — a museum of bioengineered organisms. He talks about the joys and pitfalls involved in collecting genetically modified maize, mosquitoes and zebrafish."

"Just as the financial system has to be changed to serve the needs of the people before the needs of the bankers, the drug trial system has to be changed to lower the incentives for cheating (and massive death tolls) just for a quick buck. As I mentioned before, it’s still not clear that they would have made less money, even including the penalties, if they had come clean in 2000. They made a bet that the fines they’d need to eventually pay would be smaller than the profits they’d make in the meantime. That sounds familiar to anyone who has been following the fallout from the credit crisis.
"One thing that should be changed immediately: the clinical trials for drugs should not be run or reported on by the drug companies themselves. There has to be a third party which is in charge of testing the drugs and has the power to take the drugs off the market immediately if adverse effects (like CVT events) are found. Hopefully they will be given more power than risk firms are currently given in finance (which is none)- in other words, it needs to be more than reporting, it needs to be an active regulatory power, with smart people who understand statistics and do their own state-of-the-art analyses – although as we’ve seen above even just Stats 101 would sometimes do the trick."

"Researchers have looked at comparisons between medical epidemiological research and psychological research using effect size r in an effort to compare relative effects. Often the outcomes of such efforts have demonstrated comparatively low effects for medical epidemiology research in comparison with effect sizes seen in psychology. The conclusion has often been that relatively small effects seen in psychology research are as strong as those found in important epidemiological medical research. The author suggests that many of the calculated effect sizes from medical epidemiological research on which this conclusion has been based are flawed. Specifically, rather than calculating effect sizes for treatment, many results have been for a Treatment Effect × Disease Effect interaction that was irrelevant to the main study hypothesis. A technique for developing a “hypothesis-relevant” effect size r is proposed."

"A hand-written and illustrated translation by an American artist". To make this multiply bizarre, the visual style resembles nothing so much as a medieval Book of Hours, only with scenes of contemporary American life. It seems really strange, though the pictures attached to the article aren't much on which to evaluate it.

"Our concern is selecting the concentration matrix's nonzero coefficients for a sparse Gaussian graphical model in a high-dimensional setting. This corresponds to estimating the graph of conditional dependencies between the variables. We describe a novel framework taking into account a latent structure on the concentration matrix. This latent structure is used to drive a penalty matrix and thus to recover a graphical model with a constrained topology. Our method uses an $ell_1$ penalized likelihood criterion. Inference of the graph of conditional dependencies between the variates and of the hidden variables is performed simultaneously in an iterative textsc{em}-like algorithm. The performances of our method is illustrated on synthetic as well as real data, the latter concerning breast cancer."

"For some well-known games, such as the Traveler's Dilemma or the Centipede Game, traditional game-theoretic solution concepts--and most notably Nash equilibrium--predict outcomes that are not consistent with empirical observations. In this paper, we introduce a new solution concept, iterated regret minimization, which exhibits the same qualitative behavior as that observed in experiments in many games of interest, including Traveler's Dilemma, the Centipede Game, Nash bargaining, and Bertrand competition. As the name suggests, iterated regret minimization involves the iterated deletion of strategies that do not minimize regret."

--- Quite astonishingly, no mention at all of low-regret learning!

"The paper is devoted to the problem of establishing right-convergence of sparse random graphs. This concerns the convergence of the logarithm of number of homomorphisms from graphs or hyper-graphs $G_N, Nge 1$ to some target graph $W$. The theory of dense graph convergence, including random dense graphs, is now well understood, but its counterpart for sparse random graphs presents some fundamental difficulties. Phrased in the statistical physics terminology, the issue is the existence of the log-partition function limits, also known as free energy limits, appropriately normalized for the Gibbs distribution associated with $W$. In this paper we prove that the sequence of sparse ER graphs is right-converging when the tensor product associated with the target graph $W$ satisfies certain convexity property. We treat the case of discrete and continuous target graphs $W$. The latter case allows us to prove a special case of Talagrand's recent conjecture (more accurately stated as level III Research Problem 6.7.2 in his recent book), concerning the existence of the limit of the measure of a set obtained from $R^N$ by intersecting it with linearly in $N$ many subsets, generated according to some common probability law.
Our proof is based on the interpolation technique, introduced first by Guerra and Toninelli and developed further in a series of papers. Specifically, Bayati et al establish the right-convergence property for Erdos-Renyi graphs for some special cases of $W$. In this paper most of the results in this paper follow as a special case of our main theorem."

"Computing smoothing distributions, the distributions of one or more states conditional on past, present, and future observations is a recurring problem when operating on general hidden Markov models. The aim of this paper is to provide a foundation of particle-based approximation of such distributions and to analyze, in a common unifying framework, different schemes producing such approximations. In this setting, general convergence results, including exponential deviation inequalities and central limit theorems, are established. In particular, time uniform bounds on the marginal smoothing error are obtained under appropriate mixing conditions on the transition kernel of the latent chain. In addition, we propose an algorithm approximating the joint smoothing distribution at a cost that grows only linearly with the number of particles."

"We address the online linear optimization problem with bandit feedback. Our contribution is twofold. First, we provide an algorithm (based on exponential weights) with a regret of order $sqrt{d n log N}$ for any finite action set with $N$ actions, under the assumption that the instantaneous loss is bounded by 1. This shaves off an extraneous $sqrt{d}$ factor compared to previous works, and gives a regret bound of order $d sqrt{n log n}$ for any compact set of actions. Without further assumptions on the action set, this last bound is minimax optimal up to a logarithmic factor. Interestingly, our result also shows that the minimax regret for bandit linear optimization with expert advice in $d$ dimension is the same as for the basic $d$-armed bandit with expert advice. Our second contribution is to show how to use the Mirror Descent algorithm to obtain computationally efficient strategies with minimax optimal regret bounds in specific examples. More precisely we study two canonical action sets: the hypercube and the Euclidean ball. In the former case, we obtain the first computationally efficient algorithm with a $d sqrt{n}$ regret, thus improving by a factor $sqrt{d log n}$ over the best known result for a computationally efficient algorithm. In the latter case, our approach gives the first algorithm with a $sqrt{d n log n}$ regret, again shaving off an extraneous $sqrt{d}$ compared to previous works."

Not sure how useful this is as an actual visualization, but very nice as eye candy. (And, if you look at the department totals, as an illustration of "an insurance company with an army".)

"Not to belabor the point -- the story, which purports to be set in 2511, is actually set in roughly 1985, i think. And why did this not bother me while I was reading it, only to make me angry on the bicycle, later? Because I grew up reading SF stories written before 1985. I grew up reading rediscovered-lost-colony-FTL stories in which the protagonists got lost in the woods, and it was fun. It didn't occur to me then that they would have GPS cell phones. It was easy, this morning, to simply forget the world of today, and read as if I was in 1985. But on some level this is morally bankrupt. When you don't know something, you are innocent of it. Once you do know it, though, all that is possible is feigned innocence, or incoherence."

" if some policy increases consumption of something with zero marginal cost (an idea, an academic paper, a song, an e-book, etc.), a minimum, necessary condition to restrict that policy is that the variety of affected new goods must decrease. So if music piracy increases the number of songs consumed (and the number of songs illegally downloaded in any period of time is currently much higher than worldwide sales during that period), a minimum economic justification for a government crackdown on piracy is that the number of new songs created has decreased (in this case, they have not). Applying The First Law to open access mandates, a minimum economic justification for opposing such mandates is that either open access has no benefits, or that open access will make peer reviewed journals economically infeasible."

"Haavelmo was the first to recognize the capacity of economic models to guide poli- cies. This paper describes some of the barriers that Haavelmo’s ideas have had (and still have) to overcome, and lays out a logical framework for capturing the relationships between theory, data and policy questions. The mathematical tools that emerge from this framework now enable investigators to answer complex policy and counterfactual questions using embarrassingly simple routines, some by mere inspection of the model’s structure. Several such problems are illustrated by examples, including misspecification tests, identification, mediation and introspection."

"We examine the importance of what we term ‘Fisher dynamics’- the mechanical effects of changes in interest rates, growth rates and inflation rates on debt levels independent of borrowing -for the evolution of household debt in the U.S. over a long time horizon (1929- 2011). Adapting a standard decomposition of public debt to household sector debt, we show that these factors have been important in explaining rising debt levels, especially in the past thirty years. We identify and describe three broad regimes in the growth of household debt and several shorter episodes, distinguished by the distinct roles played Fisher dynamics and borrowing behavior in the evolution of household debt. We then provide some counterfactual trajectories of debt burdens that suggest how important financial changes beginning around 1980 have been in contributing to household debt, independent of any changes in household behavior. Specifically, if average rates of growth, inflation and interest remained the same after 1980 as before 1980, household debt burdens in 2011 would have been roughly the same as they were in the early 1950s, despite the sharp increase in borrowing in the early 2000s. We then discuss the difficulties involved in deleveraging. Under scenarios involving even substantial reductions in household expenditure, returning to debt levels of the 1980s could take decades. If lower private leverage is a condition of acceptable growth,then in the absence of a substantial fall in interest rates relative to growth rates, large-scale debt forgiveness of some form may be unavoidable."

"There are two huge holes in the textbook argument.  The first is that it overlooks neighborhood effects—literally.  The most compelling argument for rent control is neighborhood stabilization, the idea that social capital in an urban environment requires stable residence patterns.  If prices are volatile, and this leads to a lot of residential turnover, the result can be a less desirable neighborhood for everyone.  Thus the quality-adjusted supply curve is partly a function of price (or at least price stability in a dynamic model), and the S and D curves are not independent of each other.  You’ll notice that not a single textbook treatment of rent control mentions stabilization as an objective, even though this is a standard element in the real-world rhetoric surrounding this issue.  Again, I’m not taking a position, just saying that the representation you get at the introductory level is an ideological construct, not an honest analysis.

"The second hole is that rent control ordinances are normally replete with measures intended to maintain supply incentives, like price increases tied to investment in housing quality or simply spreading out increases over a longer time so tenants are able to adjust.  Again, these measures may succeed or fail, but a simple horizontal line in a one-period S&D model doesn’t begin to address them.

"In fact, advocates for rent control have taken Econ 101 (most of them), but they just disagree on how large the positive and negative impacts are. The purpose of economics should be to help us think clearly about the matter—for instance by identifying the potential empirical data that could adjudicate between competing arguments—but in its textbook form it is a ritualistic way of curtailing thought."

"One of their projects is to make glowing sushi from GloFish, commercially available zebrafish that have been modified with jellyfish and coral genes to make them fluoresce. The center also filmed the Glowing Sushi Cooking Show to teach you how to make your own transgenic sushi, provided you're not afraid of eating genetically modified organisms you bought from a pet store. But they are serious when they call their signature maki the "Not in California" roll. You can't legally purchase GloFish in California. If you're anywhere else in the US, though, you can go to town on raw fluorescent zebrafish."
Yet again, I have grown up and am living in a Bruce Sterling novel.

"We study two measures of the complexity of heterogeneous extended systems, taking random Boolean networks as prototypical cases. A measure defined by Shalizi et al. for cellular automata, based on a criterion for optimal statistical prediction [1], does not distinguish between the spatial inhomogeneity of the ordered phase and the dynamical inhomogeneity of the disordered phase. A modification in which complexities of individual nodes are calculated yields vanishing complexity values for networks in the ordered and critical regimes and for highly disordered networks, peaking somewhere in the disordered regime. Individual nodes with high complexity are the ones that pass the most information from the past to the future, a quantity that depends in a nontrivial way on both the Boolean function of a given node and its location within the network."

"We describe Information Forests, an approach to classification that generalizes Random Forests by replacing the splitting criterion of non-leaf nodes from a discriminative one -- based on the entropy of the label distribution -- to a generative one -- based on maximizing the information divergence between the class-conditional distributions in the resulting partitions. The basic idea consists of deferring classification until a measure of "classification confidence" is sufficiently high, and instead breaking down the data so as to maximize this measure. In an alternative interpretation, Information Forests attempt to partition the data into subsets that are "as informative as possible" for the purpose of the task, which is to classify the data. Classification confidence, or informative content of the subsets, is quantified by the Information Divergence. Our approach relates to active learning, semi-supervised learning, mixed generative/discriminative learning."

After reading: meh.

"We present a unifying framework for information theoretic feature selection, bringing almost two decades of research on heuristic filter criteria under a single theoretical interpretation. This is in response to the question: "what are the implicit statistical assumptions of feature selection criteria based on mutual information?". To answer this, we adopt a different strategy than is usual in the feature selection literature−instead of trying to define a criterion, we derive one, directly from a clearly specified objective function: the conditional likelihood of the training labels. While many hand-designed heuristic criteria try to optimize a definition of feature 'relevancy' and 'redundancy', our approach leads to a probabilistic framework which naturally incorporates these concepts. As a result we can unify the numerous criteria published over the last two decades, and show them to be low-order approximations to the exact (but intractable) optimisation problem. The primary contribution is to show that common heuristics for information based feature selection (including Markov Blanket algorithms as a special case) are approximate iterative maximisers of the conditional likelihood. A large empirical study provides strong evidence to favour certain classes of criteria, in particular those that balance the relative size of the relevancy/redundancy terms. Overall we conclude that the JMI criterion (Yang and Moody, 1999; Meyer et al., 2008) provides the best tradeoff in terms of accuracy, stability, and flexibility with small data samples."

"A new family of tree models is proposed, which we call "differential trees." A differential tree model is constructed from multiple data sets and aims to detect distributional differences between them. The new methodology differs from the existing difference and change detection techniques in its nonparametric nature, model construction from multiple data sets, and applicability to high-dimensional data. Through a detailed study of an arson case in New Zealand, where an individual is known to have been laying vegetation fires within a certain time period, we illustrate how these models can help detect changes in the frequencies of event occurrences and uncover unusual clusters of events in a complex environment."

--- After reading, I think their exposition is needlessly hard to follow, but let me take a stab at it. In an ordinary classification tree, we are interested in the distribution of the class labels Y given the predictors X, i.e., Pr(Y|X), and make splits on X so that (in essence) the conditional entropy H[Y|X] becomes small. This is of course equivalent to making splits so that the divergence of Pr(Y|X) from Pr(Y) is maximized. What they are interested in is not classification but _describing_ how the different classes are distinct, so the relevant distribution is Pr(X|Y), and they want a big divergence between Pr(X) and Pr(X|Y).

"Served by the refrigerator’s dose and university, eisenhower favored him if there was any television he could give him in his bowls."

"I also worry that it is too easy to find correlations of this kind, and we don't have any idea just how easy until a concerted effort has been made to show that the spurious ones are not supportable. For example, if we took "has (vs. does not have) pharyngeal consonants", or "uses (vs. does not use) close front rounded vowels", would we find correlations there too? I have some colleagues here at the University of Edinburgh, within Simon Kirby's research group, who have run some informal experiments on the data Chen uses to see if dredging up spurious correlations of this kind is easy or hard, and so far they have found it jaw-droppingly easy. (I won't say any more, because I am in the weird position of producing unrefereed telegraphing of unrefereed and informal objections to an unrefereed and unpublished working paper, and it's all getting a bit too weird for me.)"

How many languages are there in Europe? Order of 10^2. How many variables can an economist get cross-country data on? Again, order of 10^2. How many discriminable syntactic features do languages have? Easily order of 10^3 if not come. Conclusion: this is not what I mean when I say that economists should do more data-mining.

I wonder what Stefan Helmreich could do with these as samples.

" But it’s not a philosophical dispute. It’s a simple case of her making up false claims based on extremely elementary errors.
And this is why I am forced to be so mean. There are just a lot of people out there exerting significant influence over the political debate who are totally unqualified. The dilemma is especially acute in the political economic field, where wealthy right-wingers have pumped so much money to subsidize the field of pro-rich people polemics that the demand for competent defenders of letting rich people keep as much of their money as possible vastly outstrips the supply. Hence the intellectual marketplace for arguments that we should tax rich people less is glutted with hackery. "

"To understand what Murray does in Coming Apart, imagine this analogy: A social scientist visits a Gulf Coast town. He notices that the houses near the water have all been smashed and shattered. The former occupants now live in tents and FEMA trailers. The social scientist writes a report: 'The evidence strongly shows that living in houses is better for children and families than living in tents and trailers. The people on the waterfront are irresponsibly subjecting their children to unacceptable conditions.'
"When he publishes his report, somebody points out: "You know, there was a hurricane here last week." The social scientist shrugs off the criticism with the reply, "I'm writing about housing, not weather." "

---All parts of Frum's review are worth reading.

"Cooperation and the allocation of common resources are core features of social behavior. Games idealizing both interactions have been studied separately. But here, rather than examining the dynamics of the individual games, the interactions are combined so that players first choose whether to cooperate, and then, if they jointly cooperate, they bargain over the fruits of their cooperation. It is shown that the dynamics of the combined game cannot simply be reduced to the dynamics of the individual games and that both cooperation and fair division are more likely in the combined game than in the constituent games taken separately."

"This article examines a series of Schelling-like models of residential segregation, in which agents prefer to be in the minority. We demonstrate that as long as agents care about the characteristics of their wider community, they tend to end up in a segregated state. We then investigate the process that causes this and conclude that the result hinges on the similarity of informational states among agents of the same type. This is quite different from Schelling-like behavior and suggests (in his terms) that segregation is an instance of macrobehavior that can arise from a wide variety of micromotives."

We knew nothing.

And here I was worried from the headline that I might have to call out Uncle Paul.

I smell an ADA take-home exam.

Since this is publicly online now, I guess I can write that post.

"There's a special place in hell for the perpetrators of this, where I hope the gods of mathematics and reason are devising some exquisite tortures for them—perhaps in the form of endlessly reading Introduction to Statistics and doing the same problem sets over and over through eternity . . ."

Balancing selection?

"A density ratio is defined by the ratio of two probability densities. We study the inference problem of density ratios and apply a semiparametric density-ratio estimator to the two-sample homogeneity test. In the proposed test procedure, the $f$-divergence between two probability densities is estimated using a density-ratio estimator. The $f$ -divergence estimator is then exploited for the two-sample homogeneity test. We derive an optimal estimator of $f$-divergence in the sense of the asymptotic variance in a semiparametric setting, and provide a statistic for two-sample homogeneity test based on the optimal estimator. We prove that the proposed test dominates the existing empirical likelihood score test. Through numerical studies, we illustrate the adequacy of the asymptotic theory for finite-sample inference."

"Static forecasting of stationary and ergodic time series is considered, i.e., inference of the conditional expectation of the response variable at time zero given the infinite past. It is shown that the mean squared error of a combination of suitably defined localized least squares estimates converges to zero for all distributions where the response variable is square integrable."

"We design an online algorithm for Principal Component Analysis. In each trial the current instance is centered and projected into a probabilistically chosen low dimensional subspace. The regret of our online algorithm, that is, the total expected quadratic compression loss of the online algorithm minus the total quadratic compression loss of the batch algorithm, is bounded by a term whose dependence on the dimension of the instances is only logarithmic.
"We first develop our methodology in the expert setting of online learning by giving an algorithm for learning as well as the best subset of experts of a certain size. This algorithm is then lifted to the matrix setting where the subsets of experts correspond to subspaces. The algorithm represents the uncertainty over the best subspace as a density matrix whose eigenvalues are bounded. The running time is O(n2) per trial, where n is the dimension of the instances."

"Recent results about the robustness of kernel methods involve the analysis of influence functions. By definition the influence function is closely related to leave-one-out criteria. In statistical learning, the latter is often used to assess the generalization of a method. In statistics, the influence function is used in a similar way to analyze the statistical efficiency of a method. Links between both worlds are explored. The influence function is related to the first term of a Taylor expansion. Higher order influence functions are calculated. A recursive relation between these terms is found characterizing the full Taylor expansion. It is shown how to evaluate influence functions at a specific sample distribution to obtain an approximation of the leave-one-out error. A specific implementation is proposed using a L1 loss in the selection of the hyperparameters and a Huber loss in the estimation procedure. The parameter in the Huber loss controlling the degree of robustness is optimized as well. The resulting procedure gives good results, even when outliers are present in the data."

"Statistical models on full and partial rankings of n items are often of limited practical use for large n due to computational consideration. We explore the use of non-parametric models for partially ranked data and derive computationally efficient procedures for their use for large n. The derivations are largely possible through combinatorial and algebraic manipulations based on the lattice of partial rankings. A bias-variance analysis and an experimental study demonstrate the applicability of the proposed method."

"Virtually all aspects of human behavior show enormous variation both within and between cultural groups, including material culture, social organization and language. Thousands of distinct cultural groups exist: about 6,000 languages are spoken today, and it is thought that a far greater number of languages existed in the past but became extinct. Using a Darwinian approach, this book seeks to explain this rich cultural variation. There are a number of theoretical reasons to believe that cultural diversification might be tree-like, that is phylogenetic: material and non-material culture is clearly inherited by descendants, there is descent with modification, and languages appear to be hierarchically related. There are also a number of theoretical reasons to believe that cultural evolution is not tree-like: cultural inheritance is not Mendelian and can indeed be vertical, horizontal or oblique, evidence of borrowing abounds, cultures are not necessarily biological populations and can be transient and complex. Here, for the first time, this title tackles these questions of cultural evolution empirically and quantitatively, using a range of case studies from Africa, the Pacific, Europe, Asia and America. A range of powerful theoretical tools developed in evolutionary biology is used to test detailed hypotheses about historical patterns and adaptive functions in cultural evolution. Evidence is amassed from archaeological, linguist and cultural datasets, from both recent and historical or pre-historical time periods. A unifying theme is that the phylogenetic approach is a useful and powerful framework, both for describing the evolutionary history of these traits, and also for testing adaptive hypotheses about their evolution and co-evolution. Contributors include archaeologists, anthropologists, evolutionary biologists and linguists, and this book will be of great interest to all those involved in these areas."

"The evolutionary history of species is traditionally represented using a rooted phylogenetic tree. However, when reticulate events such as hybridization, horizontal gene transfer or recombination are believed to be involved, phylogenetic networks that can accommodate non-treelike evolution have an important role to play. This book provides the first interdisciplinary overview of phylogenetic networks. Beginning with a concise introduction to both phylogenetic trees and phylogenetic networks, the fundamental concepts and results are then presented for both rooted and unrooted phylogenetic networks. Current approaches and algorithms available for computing phylogenetic networks from different types of datasets are then discussed, accompanied by examples of their application to real biological datasets. The book also summarises the algorithms used for drawing phylogenetic networks, along with the existing software for their computation and evaluation. All datasets, examples and other additional information and links are available from the book's companion website at www.phylogenetic-networks.org."

That was not one of Noah's better performances, no.

I hope they're not in trouble, but --- seriously, reality? This cheap imitation of eighty-year-old fiction is the best you can do?

Even if this were a complete travesty of the Greek text (and how would I know?), I am pretty sure that the book would be awesome.

"a new approach to the epistemological and methodological doctrines of the leading experimental philosophers of seventeenth-century England, an approach that considers their often overlooked moral, psychological, and theological elements. Corneanu focuses on the views about the pursuit of knowledge in the writings of Robert Boyle and John Locke, as well as in those of several of their influences, including Francis Bacon and the early Royal Society virtuosi. She argues that their experimental programs of inquiry fulfill the role of regimens for curing, ordering, and educating the mind toward an ethical purpose, an idea she tracks back to the ancient tradition of cultura animi. "

" tells the fascinating story of the transformation of cetaceans from grotesque monsters, useful only as wallowing kegs of fat and fertilizer, to playful friends of humanity, bellwethers of environmental devastation, and, finally, totems of the counterculture in the Age of Aquarius. "

Massively-hyped paper trying to model language history using lightly-repurposed biological models comprehensively debunked by very careful and linguistically-informed data analysis. One of the authors of the debunking shows up in the comments, and says:
"Finally, regarding press; a few news organisations were interested in the initial pitch, but lost interest when they realised that we didn't have a good story here about human origins."

"As I've written before, the function of the stock market in modern capitalism is to get money out of corporations, not put money into them. The social problem they are solving is not society's need to allocate scarce savings to the most promising investments, but wealth-owners desire to free their fortunes from particular firm or industry and keep them as claims on the social product as a whole."

"K-means clustering is a widely used tool for cluster analysis due to its conceptual simplicity and computational efficiency. However, its performance can be distorted when clustering high-dimensional data where the number of variables becomes relatively large and many of them may contain no information about the clustering structure. This article proposes a high-dimensional cluster analysis method via regularized k-means clustering, which can simultaneously cluster similar observations and eliminate redundant variables. The key idea is to formulate the k-means clustering in a form of regularization, with an adaptive group lasso penalty term on cluster centers. In order to optimally balance the trade-off between the clustering model fitting and sparsity, a selection criterion based on clustering stability is developed. The asymptotic estimation and selection consistency of the regularized k-means clustering with diverging dimension is established. The effectiveness of the regularized k-means clustering is also demonstrated through a variety of numerical experiments as well as applications to two gene microarray examples. The regularized clustering framework can also be extended to the general model-based clustering."