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
.
High tumor mutation burden (TMB-H) has been proposed as a predictive biomarker for response to immune checkpoint blockade (ICB), largely due to the potential for tumor mutations to generate ...immunogenic neoantigens. Despite recent pan-cancer approval of ICB treatment for any TMB-H tumor, as assessed by the targeted FoundationOne CDx assay in nine tumor types, the utility of this biomarker has not been fully demonstrated across all cancers.
Data from over 10 000 patient tumors included in The Cancer Genome Atlas were used to compare approaches to determine TMB and identify the correlation between predicted neoantigen load and CD8 T cells. Association of TMB with ICB treatment outcomes was analyzed by both objective response rates (ORRs, N = 1551) and overall survival (OS, N = 1936).
In cancer types where CD8 T-cell levels positively correlated with neoantigen load, such as melanoma, lung, and bladder cancers, TMB-H tumors exhibited a 39.8% ORR to ICB 95% confidence interval (CI) 34.9-44.8, which was significantly higher than that observed in low TMB (TMB-L) tumors odds ratio (OR) = 4.1, 95% CI 2.9-5.8, P < 2 × 10−16. In cancer types that showed no relationship between CD8 T-cell levels and neoantigen load, such as breast cancer, prostate cancer, and glioma, TMB-H tumors failed to achieve a 20% ORR (ORR = 15.3%, 95% CI 9.2-23.4, P = 0.95), and exhibited a significantly lower ORR relative to TMB-L tumors (OR = 0.46, 95% CI 0.24-0.88, P = 0.02). Bulk ORRs were not significantly different between the two categories of tumors (P = 0.10) for patient cohorts assessed. Equivalent results were obtained by analyzing OS and by treating TMB as a continuous variable.
Our analysis failed to support application of TMB-H as a biomarker for treatment with ICB in all solid cancer types. Further tumor type-specific studies are warranted.
•TMB-H failed to predict improved or clinically relevant response to ICB in all cancer types.•Cancer types where TMB-H does not predict response generally show no relationship between tumor neoantigen load and CD8 T-cell infiltration.•Further studies should be carried out before application of TMB-H as a biomarker for ICB in all cancer types.
Kuroshio velocity structure and transport in the East China Sea (ECS) were investigated as part of a 23‐month study using inverted echo sounders and acoustic Doppler current profilers (ADCPs) along ...the regularly sampled PN‐line. Flow toward the northeast is concentrated near the continental shelf with the mean surface velocity maximum located 30 km offshore from the shelf break (taken as the 170 m isobath). There are two regions of southwestward flow: a deep countercurrent over the continental slope beneath the Kuroshio axis and a recirculation offshore which extends throughout the whole water column. There is a bimodal distribution to the depth of maximum velocity with occurrence peaks at the surface and 210 dbar. When the maximum velocity is located within the top 80 m of the water column, it ranges between 0.36 m/s and 2.02 m/s; when the maximum velocity is deeper than 80 m, it ranges between 0.31 m/s and 1.11 m/s. The 13‐month mean net absolute transport of the Kuroshio in the ECS is 18.5 ± 0.8 Sv (standard deviation, σ = 4.0 Sv). The mean positive and negative portions of this net flow are 24.0 ± 0.9 Sv and −5.4 ± 0.3 Sv, respectively.
To determine if an increase in knee extensor strength mediates the effect of a 12-week knee extensor strength training program on pain and physical function improvement in people with knee ...osteoarthritis (OA).
Secondary analysis from a randomised controlled trial comparing the effects of a 12-week knee extensor strengthening exercise program to a control group with no intervention.
Data from participants with complete data (n = 97) enrolled in a previous clinical trial were analysed. Baseline and 12-week follow-up assessments included peak isometric knee extensor strength, pain and physical function. Peak knee extensor strength (Nm/kg) was assessed on an isokinetic dynamometer and subscales of the Western Ontario and McMaster Universities (WOMAC) Osteoarthritis Index were used to assess pain and physical function. Twelve-week change in pain and physical function were regressed separately, on 12-week change in knee extensor strength and group allocation. Covariates included baseline pain or physical function as appropriate, and baseline knee extensor strength, age, sex and knee alignment (stratification variable).
Improved knee extensor strength mediated the effect of the strengthening program on both pain relief (mediated effect size = 0.69, 95% confidence intervals (CI) 0.05–1.33, P = 0.03), and improved physical function (mediated effect size = 1.86, 95% CI 0.08–3.64, P = 0.04), at 12 weeks.
Increased knee extensor strength partially mediates the effect of a knee extensor strength training program on pain and physical function improvement in people with knee OA.