The portfolio optimization model has limited impact in practice because of estimation issues when applied to real data. To address this, we adapt two machine learning methods, regularization and ...cross-validation, for portfolio optimization. First, we introduce
performance-based regularization
(PBR), where the idea is to constrain the sample variances of the estimated portfolio risk and return, which steers the solution toward one associated with less estimation error in the performance. We consider PBR for both mean-variance and mean-conditional value-at-risk (CVaR) problems. For the mean-variance problem, PBR introduces a quartic polynomial constraint, for which we make two convex approximations: one based on rank-1 approximation and another based on a convex quadratic approximation. The rank-1 approximation PBR adds a bias to the optimal allocation, and the convex quadratic approximation PBR shrinks the sample covariance matrix. For the mean-CVaR problem, the PBR model is a combinatorial optimization problem, but we prove its convex relaxation, a quadratically constrained quadratic program, is essentially tight. We show that the PBR models can be cast as robust optimization problems with novel uncertainty sets and establish asymptotic optimality of both sample average approximation (SAA) and PBR solutions and the corresponding efficient frontiers. To calibrate the right-hand sides of the PBR constraints, we develop new, performance-based
k
-fold cross-validation algorithms. Using these algorithms, we carry out an extensive empirical investigation of PBR against SAA, as well as L1 and L2 regularizations and the equally weighted portfolio. We find that PBR dominates all other benchmarks for two out of three Fama–French data sets.
This paper was accepted by Yinyu Ye, optimization
.
Cross-sectional studies suggest that sleep fragmentation is associated with cognitive performance in older adults. We tested the hypothesis that sleep fragmentation is associated with incident ...Alzheimer's disease (AD) and the rate of cognitive decline in older adults.
Prospective cohort study.
Community-based.
737 community dwelling older adults without dementia.
Sleep fragmentation was quantified from up to 10 consecutive days of actigraphy. Subjects underwent annual evaluation for AD with 19 neuropsychological tests. Over a follow-up period of up to 6 years (mean 3.3 years), 97 individuals developed AD. In a Cox proportional hazards model controlling for age, sex, and education, a higher level of sleep fragmentation was associated with an increased risk of AD (HR = 1.22, 95%CI 1.03-1.44, P = 0.02 per 1SD increase in sleep fragmentation). An individual with high sleep fragmentation (90
percentile) had a 1.5-fold risk of developing AD as compared with someone with low sleep fragmentation (10
percentile). The association of sleep fragmentation with incident AD did not vary along demographic lines and was unchanged after controlling for potential confounders including total daily rest time, chronic medical conditions, and the use of common medications which can affect sleep. In a linear mixed effect analysis, a 0.01 unit increase in sleep fragmentation was associated with a 22% increase in the annual rate of cognitive decline relative to the average rate of decline in the cohort (Estimate = -0.016, SE = 0.007, P = 0.03).
Sleep fragmentation in older adults is associated with incident AD and the rate of cognitive decline.
Lim ASP; Kowgier M; Yu L; Buchman AS; Bennett DA. Sleep fragmentation and the risk of incident alzheimer's disease and cognitive decline in older persons.
2013;36(7):1027-1032.
IMPORTANCE The apolipoprotein E (APOE GenBank, 348; OMIM, 107741) ε4 allele is a common and well-established genetic risk factor for Alzheimer disease (AD). Sleep consolidation is also associated ...with AD risk, and previous work suggests that APOE genotype and sleep may interact to influence cognitive function. OBJECTIVE To determine whether better sleep consolidation attenuates the relationship of the APOE genotype to the risk of incident AD and the burden of AD pathology. DESIGN, SETTING, AND PARTICIPANTS A prospective longitudinal cohort study with up to 6 years of follow-up was conducted. Participants included 698 community-dwelling older adults without dementia (mean age, 81.7 years; 77% women) in the Rush Memory and Aging Project. EXPOSURES We used up to 10 days of actigraphic recording to quantify the degree of sleep consolidation and ascertained APOE genotype. MAIN OUTCOMES AND MEASURES Participants underwent annual evaluation for AD during a follow-up period of up to 6 years. Autopsies were performed on 201 participants who died, and β-amyloid (Aβ) and neurofibrillary tangles were identified by immunohistochemistry and quantified. RESULTS During the follow-up period, 98 individuals developed AD. In a series of Cox proportional hazards regression models, better sleep consolidation attenuated the effect of the ε4 allele on the risk of incident AD (hazard ratio, 0.67; 95% CI, 0.46-0.97; P = .04 per allele per 1-SD increase in sleep consolidation). In a series of linear mixed-effect models, better sleep consolidation also attenuated the effect of the ε4 allele on the annual rate of cognitive decline. In individuals who died, better sleep consolidation attenuated the effect of the ε4 allele on neurofibrillary tangle density (interaction estimate, −0.42; SE = 0.17; P = .02), which accounted for the effect of sleep consolidation on the association between APOE genotype and cognition proximate to death. CONCLUSIONS AND RELEVANCE Better sleep consolidation attenuates the effect of APOE genotype on incident AD and development of neurofibrillary tangle pathology. Assessment of sleep consolidation may identify APOE+ individuals at high risk for incident AD, and interventions to enhance sleep consolidation should be studied as potentially useful means to reduce the risk of AD and development of neurofibrillary tangles in APOE ε4+ individuals.
Sleep problems are both symptoms of and modifiable risk factors for many psychiatric disorders. Wrist-worn accelerometers enable objective measurement of sleep at scale. Here, we aimed to examine the ...association of accelerometer-derived sleep measures with psychiatric diagnoses and polygenic risk scores in a large community-based cohort.
In this post hoc cross-sectional analysis of the UK Biobank cohort, 10 interpretable sleep measures-bedtime, wake-up time, sleep duration, wake after sleep onset, sleep efficiency, number of awakenings, duration of longest sleep bout, number of naps, and variability in bedtime and sleep duration-were derived from 7-day accelerometry recordings across 89,205 participants (aged 43 to 79, 56% female, 97% self-reported white) taken between 2013 and 2015. These measures were examined for association with lifetime inpatient diagnoses of major depressive disorder, anxiety disorders, bipolar disorder/mania, and schizophrenia spectrum disorders from any time before the date of accelerometry, as well as polygenic risk scores for major depression, bipolar disorder, and schizophrenia. Covariates consisted of age and season at the time of the accelerometry recording, sex, Townsend deprivation index (an indicator of socioeconomic status), and the top 10 genotype principal components. We found that sleep pattern differences were ubiquitous across diagnoses: each diagnosis was associated with a median of 8.5 of the 10 accelerometer-derived sleep measures, with measures of sleep quality (for instance, sleep efficiency) generally more affected than mere sleep duration. Effect sizes were generally small: for instance, the largest magnitude effect size across the 4 diagnoses was β = -0.11 (95% confidence interval -0.13 to -0.10, p = 3 × 10-56, FDR = 6 × 10-55) for the association between lifetime inpatient major depressive disorder diagnosis and sleep efficiency. Associations largely replicated across ancestries and sexes, and accelerometry-derived measures were concordant with self-reported sleep properties. Limitations include the use of accelerometer-based sleep measurement and the time lag between psychiatric diagnoses and accelerometry.
In this study, we observed that sleep pattern differences are a transdiagnostic feature of individuals with lifetime mental illness, suggesting that they should be considered regardless of diagnosis. Accelerometry provides a scalable way to objectively measure sleep properties in psychiatric clinical research and practice, even across tens of thousands of individuals.
Celotno besedilo
Dostopno za:
DOBA, IZUM, KILJ, NUK, PILJ, PNG, SAZU, SIK, UILJ, UKNU, UL, UM, UPUK
Robust Multiarmed Bandit Problems Kim, Michael Jong; Lim, Andrew E.B.
Management science,
01/2016, Letnik:
62, Številka:
1
Journal Article
Recenzirano
The multiarmed bandit problem is a popular framework for studying the exploration versus exploitation trade-off. Recent applications include dynamic assortment design, Internet advertising, dynamic ...pricing, and the control of queues. The standard mathematical formulation for a bandit problem makes the strong assumption that the decision maker has a full characterization of the joint distribution of the rewards, and that ″arms″ under this distribution are independent. These assumptions are not satisfied in many applications, and the outof-sample performance of policies that optimize a misspecified model can be poor. Motivated by these concerns, we formulate a robust bandit problem in which a decision maker accounts for distrust in the nominal model by solving a worst-case problem against an adversary (″nature″) who has the ability to alter the underlying reward distribution and does so to minimize the decision maker's expected total profit. Structural properties of the optimal worst-case policy are characterized by using the robust Bellman (dynamic programming) equation, and arms are shown to be no longer independent under nature's worst-case response. One implication of this is that index policies are not optimal for the robust problem, and we propose, as an alternative, a robust version of the Gittins index. Performance bounds for the robust Gittins index are derived by using structural properties of the value function together with ideas from stochastic dynamic programming duality. We also investigate the performance of the robust Gittins index policy when applied to a Bayesian webpage design problem. In the presence of model misspecification, numerical experiments show that the robust Gittins index policy not only outperforms the classical Gittins index policy, but also substantially reduces the variability in the out-of-sample performance.
There are few data concerning the association between season and cognition and its neurobiological correlates in older persons-effects with important translational and therapeutic implications for ...the diagnosis and treatment of Alzheimer disease (AD). We aimed to measure these effects.
We analyzed data from 3,353 participants from 3 observational community-based cohort studies of older persons (the Rush Memory and Aging Project MAP, the Religious Orders Study ROS, and the Minority Aging Research Study MARS) and 2 observational memory-clinic-based cohort studies (Centre de Neurologie Cognitive CNC study at Lariboisière Hospital and the Sunnybrook Dementia Study SDS). We performed neuropsychological testing and, in subsets of participants, evaluated cerebrospinal fluid AD biomarkers, standardized structured autopsy measures, and/or prefrontal cortex gene expression by RNA sequencing. We examined the association between season and these variables using nested multiple linear and logistic regression models. There was a robust association between season and cognition that was replicated in multiple cohorts (amplitude = 0.14 SD a measure of the magnitude of seasonal variation relative to overall variability; 95% CI 0.07-0.23, p = 0.007, in the combined MAP, ROS, and MARS cohorts; amplitude = 0.50 SD 95% CI 0.07-0.66, p = 0.017, in the SDS cohort). Average composite global cognitive function was higher in the summer and fall compared to winter and spring, with the difference equivalent in cognitive effect to 4.8 years' difference in age (95% CI 2.1-8.4, p = 0.002). Further, the odds of meeting criteria for mild cognitive impairment or dementia were higher in the winter and spring (odds ratio 1.31 95% CI 1.10-1.57, p = 0.003). These results were robust against multiple potential confounders including depressive symptoms, sleep, physical activity, and thyroid status and persisted in cases with AD pathology. Moreover, season had a marked effect on cerebrospinal fluid Aβ 42 level (amplitude 0.30 SD 95% CI 0.10-0.64, p = 0.003), which peaked in the summer, and on the brain expression of 4 cognition-associated modules of co-expressed genes (m6: amplitude = 0.44 SD 95% CI 0.21-0.65, p = 0.0021; m13: amplitude = 0.46 SD 95% CI 0.27-0.76, p = 0.0009; m109: amplitude = 0.43 SD 95% CI 0.24-0.67, p = 0.0021; and m122: amplitude 0.46 SD 95% CI 0.20-0.71, p = 0.0012), which were in phase or anti-phase to the rhythms of cognition and which were in turn associated with binding sites for several seasonally rhythmic transcription factors including BCL11A, CTCF, EGR1, MEF2C, and THAP1. Limitations include the evaluation of each participant or sample once per annual cycle, reliance on self-report for measurement of environmental and behavioral factors, and potentially limited generalizability to individuals in equatorial regions or in the southern hemisphere.
Season has a clinically significant association with cognition and its neurobiological correlates in older adults with and without AD pathology. There may be value in increasing dementia-related clinical resources in the winter and early spring, when symptoms are likely to be most pronounced. Moreover, the persistence of robust seasonal plasticity in cognition and its neurobiological correlates, even in the context of concomitant AD pathology, suggests that targeting environmental or behavioral drivers of seasonal cognitive plasticity, or the key transcription factors and genes identified in this study as potentially mediating these effects, may allow us to substantially improve cognition in adults with and without AD.
Celotno besedilo
Dostopno za:
DOBA, IZUM, KILJ, NUK, PILJ, PNG, SAZU, SIK, UILJ, UKNU, UL, UM, UPUK
Fragmented sleep is a common and troubling symptom in ageing and Alzheimer's disease; however, its neurobiological basis in many patients is unknown. In rodents, lesions of the hypothalamic ...ventrolateral preoptic nucleus cause fragmented sleep. We previously proposed that the intermediate nucleus in the human hypothalamus, which has a similar location and neurotransmitter profile, is the homologue of the ventrolateral preoptic nucleus, but physiological data in humans were lacking. We hypothesized that if the intermediate nucleus is important for human sleep, then intermediate nucleus cell loss may contribute to fragmentation and loss of sleep in ageing and Alzheimer's disease. We studied 45 older adults (mean age at death 89.2 years; 71% female; 12 with Alzheimer's disease) from the Rush Memory and Aging Project, a community-based study of ageing and dementia, who had at least 1 week of wrist actigraphy proximate to death. Upon death a median of 15.5 months later, we used immunohistochemistry and stereology to quantify the number of galanin-immunoreactive intermediate nucleus neurons in each individual, and related this to ante-mortem sleep fragmentation. Individuals with Alzheimer's disease had fewer galaninergic intermediate nucleus neurons than those without (estimate -2872, standard error = 829, P = 0.001). Individuals with more galanin-immunoreactive intermediate nucleus neurons had less fragmented sleep, after adjusting for age and sex, and this association was strongest in those for whom the lag between actigraphy and death was <1 year (estimate -0.0013, standard error = 0.0005, P = 0.023). This association did not differ between individuals with and without Alzheimer's disease, and similar associations were not seen for two other cell populations near the intermediate nucleus. These data are consistent with the intermediate nucleus being the human homologue of the ventrolateral preoptic nucleus. Moreover, they demonstrate that a paucity of galanin-immunoreactive intermediate nucleus neurons is accompanied by sleep fragmentation in older adults with and without Alzheimer's disease.
Circadian rhythms modulate the biology of many human tissues, including brain tissues, and are driven by a near 24-hour transcriptional feedback loop. These rhythms are paralleled by 24-hour rhythms ...of large portions of the transcriptome. The role of dynamic DNA methylation in influencing these rhythms is uncertain. While recent work in Neurospora suggests that dynamic site-specific circadian rhythms of DNA methylation may play a role in modulating the fungal molecular clock, such rhythms and their relationship to RNA expression have not, to our knowledge, been elucidated in mammalian tissues, including human brain tissues. We hypothesized that 24-hour rhythms of DNA methylation exist in the human brain, and play a role in driving 24-hour rhythms of RNA expression. We analyzed DNA methylation levels in post-mortem human dorsolateral prefrontal cortex samples from 738 subjects. We assessed for 24-hour rhythmicity of 420,132 DNA methylation sites throughout the genome by considering methylation levels as a function of clock time of death and parameterizing these data using cosine functions. We determined global statistical significance by permutation. We then related rhythms of DNA methylation with rhythms of RNA expression determined by RNA sequencing. We found evidence of significant 24-hour rhythmicity of DNA methylation. Regions near transcription start sites were enriched for high-amplitude rhythmic DNA methylation sites, which were in turn time locked to 24-hour rhythms of RNA expression of nearby genes, with the nadir of methylation preceding peak transcript expression by 1-3 hours. Weak ante-mortem rest-activity rhythms were associated with lower amplitude DNA methylation rhythms as were older age and the presence of Alzheimer's disease. These findings support the hypothesis that 24-hour rhythms of DNA methylation, particularly near transcription start sites, may play a role in driving 24-hour rhythms of gene expression in the human dorsolateral prefrontal cortex, and may be affected by age and Alzheimer's disease.
Celotno besedilo
Dostopno za:
DOBA, IZUM, KILJ, NUK, PILJ, PNG, SAZU, SIK, UILJ, UKNU, UL, UM, UPUK
Learning about experts in a portfolio choice problem:
The Black–Litterman model provides a framework for combining the forecasts of an equilibrium model and the forward-looking opinions of several ...experts in a portfolio allocation decision. In “A Generalized Black–Litterman Model,” Chen and Lim propose a generalization of the classical model that accounts for model misspecification and bias in the equilibrium and expert models and show how it can be calibrated using historical view and return data. More generally, this paper shows how the views of multiple experts can be modeled as a Bayesian graphical model and estimated using historical data, which may be of interest in applications that involve the aggregation of expert opinions for the purpose of decision making.
The Black–Litterman model provides a framework for combining the forecasts of a backward-looking equilibrium model with the views of (several) forward-looking experts in a portfolio allocation decision. The classical version uses the capital asset pricing model to specify expected returns, and assumes that expert views are unbiased noisy observations of future returns. It combines the two using Bayes’ rule and the portfolio allocation decision is made on the basis of the updated forecast. The classical Black–Litterman model assumes that the equilibrium and expert models are accurately specified. This is generally not the case, however, and there may be substantial efficiency loss if misspecification is ignored. In this paper, we formulate a generalized Black–Litterman model that accounts for both misspecification and bias in the equilibrium and expert models. We show how to calibrate this model using historical view and return data, and study the value of our generalized model for portfolio construction. More generally, this paper shows how the views of multiple experts can be modeled as a Bayesian graphical model and estimated using historical data, which may be of interest in applications that involve the aggregation of expert opinions for the purpose of decision making.
Abstract Introduction Circadian alterations are prevalent in Alzheimer's disease (AD) and may contribute to cognitive impairment, behavioral symptoms, and neurodegeneration. Epigenetic mechanisms ...regulate the circadian clock, and changes in DNA methylation have been reported in AD brains, but the pathways that mediate circadian deregulation in AD are incompletely understood. We hypothesized that aberrant DNA methylation may affect circadian rhythms in AD. Methods We investigated DNA methylation, transcription, and expression of BMAL1 , a positive regulator of the circadian clock, in cultured fibroblasts and brain samples from two independent cohorts of aging and AD. Results DNA methylation modulated rhythmic expression of clock genes in cultured fibroblasts. Moreover, rhythmic methylation of BMAL1 was altered in AD brains and fibroblasts and correlated with transcription cycles. Discussion Our results indicate that cycles of DNA methylation contribute to the regulation of BMAL1 rhythms in the brain. Hence, aberrant epigenetic patterns may be linked to circadian alterations in AD.