Stability of feature selection in classification issues for high-dimensional correlated data

Handling dependence or not in feature selection is still an open question in supervised classification issues where the number of covariates exceeds the number of observations. Some recent papers surprisingly show the superiority of naive Bayes approaches based on an obviously erroneous assumption of independence, whereas others recommend to infer on the dependence structure in order to decorrelate the selection statistics. In the classical linear discriminant analysis (LDA) framework, the present paper first highlights the impact of dependence in terms of instability of feature selection. A second objective is to revisit the above issue using a flexible factor modeling for the covariance. This framework introduces latent components of dependence, conditionally on which a new Bayes consistency is defined. A procedure is then proposed for the joint estimation of the expectation and variance parameters of the model. The present method is compared to recent regularized diagonal discriminant analysis approaches, assuming independence among features, and regularized LDA procedures, both in terms of classification performance and stability of feature selection. The proposed method is implemented in the R package FADA, freely available from the R repository CRAN.     

European System of Accounts: The French Case

The 1995 European System of Accounts (1995 ESA), an adapted version of the United Nations System of National Accounts (1993 SNA), will be applied in most European countries. This research note shows that the empirical knowledge of the nonprofit sector is not improved in this change. We first review the treatment of the nonprofit organizations in 1968 SNA, and then discuss the 1993 SNA and 1995 ESA approaches, which do not clear up many conceptual difficulties. The classification of the private nonprofit institutions serving households by the French statistical office is also presented and criticized. A way forward through the use of a satellite account system is proposed.     

Negative Dependence in Sampling

Abstract. The strong Rayleigh property is a new and robust negative dependence property that implies negative association; in fact it implies conditional negative association closed under external fields (CNA+). Suppose that and are two families of 0‐1 random variables that satisfy the strong Rayleigh property and let . We show that {Z_i} conditioned on is also strongly Rayleigh; this turns out to be an easy consequence of the results on preservation of stability of polynomials of Borcea & Brändén (Invent. Math., 177, 2009, 521–569). This entails that a number of important π ps sampling algorithms, including Sampford sampling and Pareto sampling, are CNA+. As a consequence, statistics based on such samples automatically satisfy a version of the Central Limit Theorem for triangular arrays.     

The Role of Rhetoric in a Dialogical Approach to Thinking

The central idea of the paper is that human thinking consists in a movement through which a person socially interacts with herself. Consequently, thinking does not offer the experience of a private refuge in the intimacy of the individual thinker's self‐knowing, but a field where multiple points of view interact by contesting, distancing, approaching, agreeing or disagreeing, one to another. Classical ( Isocrates, 1929 /1968) and contemporary (Billig, 1987) rhetorical approaches to thinking stress that both “inner” and “social” discourse are addressed to someone else, are determined by the anticipation of this audience, and both are interested in persuading it. In doing so, the discursive, rhetoric, and dialogic aspects of thinking become tied to argumentation. The paper tries to show, following the dialogical notion of discourse of Bakhtin (1986) and Volo?inov (1929/1986) that, since every act of thinking consists in the raising of a point of view addressing another one and oriented by a particular interest, every stream of thought involves a rhetorical activity. A distinction between rhetoric and argumentation is proposed. On this basis, the rhetorical nature of thinking is discussed beyond argumentative discourse. Overall, this discussion contributes to a rhetorical approach to dialogism.     

THE EFFECTS OF IMMIGRATION IN A DISTORTED TWO-SECTOR ECONOMY

This paper examines the income, employment and distributional impact of immigration on a recipient economy facing an administered wage distortion. It is found that immigration increases the unemployment rate of the native population and shifts income distribution against labor and in favor of the nonlabor input. Its effect on the total income of the non-immigrants is, however, ambiguous, and depends on the relative importance of the increase in unemployment versus the net increase in the rent received by the owners of the non-labor input.     

Promoting the associational power of workers in globalized production networks: A missed opportunity

The author discusses the capacity of labour NGOs to strengthen the associational power of Chinese workers in globalized production networks by using their leverage to pressure multinationals into launching programmes to elect workers’ committees in factories. Based on an empirical survey conducted between 2012 and 2019 in southern China, he shows that such programmes helped to raise workers’ awareness of collective action and strengthen their agency, but that they ground to a halt in the face of the structural inequality of the production system, the lack of institutional protection for elected representatives and a hardening regime.