Wealth Flow Model: Online Portfolio Selection Based on Learning Wealth Flow Matrices

This article proposes a deep learning solution to the online portfolio selection problem based on learning a latent structure directly from a price time series. It introduces a novel wealth flow matrix for representing a latent structure that has special regular conditions to encode the knowledge about the relative strengths of assets in portfolios. Therefore, a wealth flow model (WFM) is proposed to learn wealth flow matrices and maximize portfolio wealth simultaneously. Compared with existing approaches, our work has several distinctive benefits: (1) the learning of wealth flow matrices makes our model more generalizable than models that only predict wealth proportion vectors, and (2) the exploitation of wealth flow matrices and the exploration of wealth growth are integrated into our deep reinforcement algorithm for the WFM. These benefits, in combination, lead to a highly-effective approach for generating reasonable investment behavior, including short-term trend following, the following of a few losers, no self-investment, and sparse portfolios. Extensive experiments on five benchmark datasets from real-world stock markets confirm the theoretical advantage of the WFM, which achieves the Pareto improvements in terms of multiple performance indicators and the steady growth of wealth over the state-of-the-art algorithms.

Locally Linear Embedding Orthogonal Projection to Latent Structure

Quality variables are measured much less frequently and usually with a significant time delay by comparison with the measurement of process variables. Monitoring process variables and their associated quality variables is essential undertaking as it can lead to potential hazards that may cause system shutdowns and thus possibly huge economic losses. Maximum correlation was extracted between quality variables and process variables by partial least squares analysis (PLS) (Kruger et al. 2001; Song et al. 2004; Li et al. 2010; Hu et al. 2013; Zhang et al. 2015).

New Robust Projection to Latent Structure

In many actual nonlinear systems, especially near the equilibrium point, linearity is the primary feature and nonlinearity is the secondary feature. For the system that deviates from the equilibrium point, the secondary nonlinearity or local structure feature can also be regarded as the small uncertainty part, just as the nonlinearity can be used to represent the uncertainty of a system (Wang et al. 2019). So this chapter also focuses on how to deal with the nonlinearity in PLS series method, but starts from an different view, i.e., robust PLS. Here the system nonlinearity is considered as uncertainty and a new robust $$\mathrm{L}_1$$ L 1 -PLS is proposed.

The Dickman Impulsivity Inventory: Validation and measurement invariance among Portuguese young adults

The Dickman Impulsivity Inventory (DII) measures impulsive personality related to both negative and positive behaviors and characteristics. The main aim of the present study was to examine the psychometric properties of the DII among a Southern-European sample of Portuguese young adults. Our convenience sample (N = 429, M = 22.11 years, SD = 3.35, range = 18–42), composed of women (n = 237, M = 22.08 years, SD = 3.35, age range = 18–42) and men (n = 192, M = 22.14 years, SD = 3.34, range = 18–35), was collected from a university context. The two-factor latent structure of the DII composed of functional and dysfunctional impulsivity was supported, although three items had to be removed due to low standardized loadings, and strong cross-gender measurement invariance was established. Our analyses of the DII also provided evidence of criterion-related validity, known-groups validity, and internal consistency/reliability. Our findings support the use of the DII among Portuguese young adults.

Unsupervised Embedding of Trajectories Captures the Latent Structure of Mobility

Human mobility drives major societal phenomena including epidemics, economies, and innovation. Historically, mobility was constrained by geographic distance, however, in the globalizing world, language, culture, and history are increasingly important. Here, we show a mathematical equivalence between word2vec model and the gravity model of mobility and demonstrate that, by using three human trajectory datasets, word2vec encodes nuanced relationships between locations into a systematic and meaningful vector-space, providing a functional distance between locations, as well as a representation for studying the many dimensions of mobility. Focusing on the case of scientific mobility, we show that embeddings implicitly learn cultural, linguistic, and hierarchical relationships at multiple levels of granularity. Connecting neural embeddings to the gravity model opens up new avenues for the study of mobility.

Impact of Chilling Rate on the Evolution of Volatile and Non-Volatile Compounds in Raw Lamb Meat during Refrigeration

The aim of this study was to investigate the effect of chilling rate (1.44, 22.2, and 32.4 °C/h) on the evolution of volatile and non-volatile compounds in raw lamb meat during refrigeration (1, 24, 72, and 120 h). Through orthogonal projection to latent structure-discriminant analysis, the calculation of odor activity values (OAV > 1) and taste activity values (TAV > 1) analysis, 1-octen-3-ol, (E, E)-2,4-decadienal, nonanal, hexanal, nona-3,5-dien-2-one, 2,3-octanedione, hexanoic acid, 1-nonen-4-ol, aspartate (Asp), Glutamic Acid (Glu), 5′-GMP, 5′-IMP, and 5′-AMP were regarded as differential flavor or taste compounds for raw meat undergone different chilling rates. With a rapid chilling rate at 24 h after slaughter, the contribution of 1-octen-3-ol decreased, but (E, E)-2,4-decadienal increased. Moreover, at 24 h post-mortem, the equivalent umami concentration of Asp, Glu, 5′-GMP, 5′-IMP and 5′-AMP in raw meat were significantly lower at a chilling rate of 1.44 °C/h than 32.4 °C/h (p < 0.05). Conclusively, under the rapid chilling rate, more fatty odor and umami compounds accumulated in 24 h aged meat.

Establishing Multivariate Specification Regions for Incoming Raw Materials Using Projection to Latent Structure Models: Comparison Between Direct Mapping and Model Inversion

Increasing raw material variability is challenging for many industries since it adversely impacts final product quality. Establishing multivariate specification regions for selecting incoming lot of raw materials is a key solution to mitigate this issue. Two data-driven approaches emerge from the literature for defining these specifications in the latent space of Projection to Latent Structure (PLS) models. The first is based on a direct mapping of good quality final product and associated lots of raw materials in the latent space, followed by selection of boundaries that minimize or best balance type I and II errors. The second rather defines specification regions by inverting the PLS model for each point lying on final product acceptance limits. The objective of this paper is to compare both methods to determine their advantages and drawbacks, and to assess their classification performance in presence of different levels of correlation between the quality attributes. The comparative analysis is performed using simulated raw materials and product quality data generated under multiple scenarios where product quality attributes have different degrees of collinearity. First, a simple case is proposed using one quality attribute to illustrate the methods. Then, the impact of collinearity is studied. It is shown that in most cases, correlation between the quality variable does not seem to influence classification performance except when the variables are highly correlated. A summary of the main advantages and disadvantages of both approaches is provided to guide the selection of the most appropriate approach for establishing multivariate specification regions for a given application.

Arguing in Favor of Revising the Simulator Sickness Questionnaire Factor Structure When Assessing Side Effects Induced by Immersions in Virtual Reality

Two issues are increasingly of interest in the scientific literature regarding unwanted virtual reality (VR) induced side effects: (1) whether the latent structure of the Simulator Sickness Questionnaire (SSQ) is comprised of two or three factors, and (2) if the SSQ measures symptoms of anxiety that can be misattributed to unwanted negative side effects induced by immersions in VR. Study 1 was conducted with a sample of 876 participants. A confirmatory factor analysis clearly supported a two-factor model composed of nausea and oculomotor symptoms instead of the 3-factor structure observed in simulators. To tease-out symptoms of anxiety from unwanted negative side effects induced by immersions in VR, Study 2 was conducted with 88 participants who were administered the Trier Stress Social Test in groups without being immersed in VR. A Spearman correlation showed that 11 out of 16 side effects correlated significantly with anxiety. A factor analysis revealed that items measuring general discomfort, difficulty concentrating, sweating, nausea, and vertigo loaded significantly on the anxiety factor comprised of items from the State-Trait Anxiety Inventory. Finally, a multiple regression indicated that the items measuring general discomfort and difficulty concentrating significantly predicted increases in anxiety. The overall results support the notion that side effects associated with immersions in VR consist mostly of a nausea and an oculomotor latent structure and that a few items are confounding anxiety and cybersickness. The data support the suggestion to revise the scoring procedures of the Simulator Sickness Questionnaire when using this instrument with immersions in VR.