Personalized medicine has received increasing attention among statisticians, computer scientists, and clinical practitioners. A major component of personalized medicine is the estimation of individualized treatment rules (ITRs). Recently, Zhao et al. (2012) proposed outcome weighted learning (OWL) to construct ITRs that directly optimize the clinical outcome. Although OWL opens the door to introducing machine learning techniques to optimal treatment regimes, it still has some problems in performance. (1) The estimated ITR of OWL is affected by a simple shift of the outcome. (2) The rule from OWL tries to keep treatment assignments that subjects actually received. (3) There is no variable selection mechanism with OWL. All of them weaken the finite sample performance of OWL. In this article, we propose a general framework, called Residual Weighted Learning (RWL), to alleviate these problems, and hence to improve finite sample performance. Unlike OWL which weights misclassification errors by clinical outcomes, RWL weights these errors by residuals of the outcome from a regression fit on clinical covariates excluding treatment assignment. We utilize the smoothed ramp loss function in RWL, and provide a difference of convex (d.c.) algorithm to solve the corresponding non-convex optimization problem. By estimating residuals with linear models or generalized linear models, RWL can effectively deal with different types of outcomes, such as continuous, binary and count outcomes. We also propose variable selection methods for linear and nonlinear rules, respectively, to further improve the performance. We show that the resulting estimator of the treatment rule is consistent. We further obtain a rate of convergence for the difference between the expected outcome using the estimated ITR and that of the optimal treatment rule. The performance of the proposed RWL methods is illustrated in simulation studies and in an analysis of cystic fibrosis clinical trial data.

ObjectiveTo evaluate the service readiness of health facilities in Bangladesh, Haiti, Kenya, Malawi, Namibia, Nepal, Rwanda, Senegal, Uganda and the United Republic of Tanzania.MethodsUsing existing data from service provision assessments of the health systems of the 10 study countries, we calculated a service readiness index for each of 8443 health facilities. This index represents the percentage availability of 50 items that the World Health Organization considers essential for providing health care. For our analysis we used 37–49 of the items on the list. We used linear regression to assess the independent explanatory power of four national and four facility-level characteristics on reported service readiness.FindingsThe mean values for the service readiness index were 77% for the 636 hospitals and 52% for the 7807 health centres/clinics. Deficiencies in medications and diagnostic capacity were particularly common. The readiness index varied more between hospitals and health centres/clinics in the same country than between them. There was weak correlation between national factors related to health financing and the readiness index.ConclusionMost health facilities in our study countries were insufficiently equipped to provide basic clinical care. If countries are to bolster health-system capacity towards achieving universal coverage, more attention needs to be given to within-country inequities.

In the present study, we propose a family of four extensions to SVM-RFE (called MSVM-RFE) to solve the multiclass gene selection problem, based on different frameworks of multiclass SVMs. By simultaneously considering all classes during the gene selection stages, our proposed extensions identify genes leading to more accurate classification.

performance and cost are not satisfactory while the energy density of conventional LIBs have almost reached the theoretical maximum. To further advance current LIBs, numerous efforts have been devoted to the exploration of new electrode and electrolyte materials. [3] The historical material research has heavily relied on either "trial-and-error" processes or serendipity, both of which require the vast numbers of tedious experiments (Figure 1a). Such intuition-based approaches are often time-consuming and inefficient, which cannot avoid the consumption of many manpower and material resources. In the past 50 years, computational chemistry, such as firstprinciples (FP) calculations, [4,5] quantum mechanics, [6] molecular dynamics (MD) [7] and Monte Carlo techniques, [8] has become a mature approach to complement and aid experimental studies for predicting and designing new materials. With the rapid development of high-performance computations, density functional theory (DFT) has been widely applied to high-throughput property prediction, which is conducive to the development of materials databases, such as Inorganic Crystal Structure Database (ICSD), [9] Cambridge Structural Database, [10] the Materials Project [11] database, AFLOWLIB consortium, [12] Open Quantum Materials Database, [13] Harvard Clean Energy Project, [14] Electronic Structure Project, [15] MaterialGo, [16] and so on. However, Lithium-ion batteries (LIBs) are vital energy-storage devices in modern society. However, the performance and cost are still not satisfactory in terms of energy density, power density, cycle life, safety, etc. To further improve the performance of batteries, traditional "trial-and-error" processes require a vast number of tedious experiments. Computational chemistry and artificial intelligence (AI) can significantly accelerate the research and development of novel battery systems. Herein, a heterogeneous category of AI technology for predicting and discovering battery materials and estimating the state of the battery system is reviewed. Successful examples, the challenges of deploying AI in real-world scenarios, and an integrated framework are analyzed and outlined. The state-of-the-art research about the applications of ML in the property prediction and battery discovery, including electrolyte and electrode materials, are further summarized. Meanwhile, the prediction of battery states is also provided. Finally, various existing challenges and the framework to tackle the challenges on the further development of machine learning for rechargeable LIBs are proposed.

Motivation:One problem with discriminant analysis of DNA microarray data is that each sample is represented by quite a large number of genes, and many of them are irrelevant, insignificant or redundant to the discriminant problem at hand. Methods for selecting important genes are, therefore, of much significance in microarray data analysis. In the present study, a new criterion, called LS Bound measure, is proposed to address the gene selection problem. The LS Bound measure is derived from leave-one-out procedure of LS-SVMs (least squares support vector machines), and as the upper bound for leave-oneout classification results it reflects to some extent the generalization performance of gene subsets. Results: We applied this LS Bound measure for gene selection on two benchmark microarray datasets: colon cancer and leukemia. We also compared the LS Bound measure with other evaluation criteria, including the well-known Fisher's ratio and Mahalanobis class separability measure, and other published gene selection algorithms, including Weighting factor and SVM Recursive Feature Elimination. The strength of the LS Bound measure is that it provides gene subsets leading to more accurate classification results than the filter method while its computational complexity is at the level of the filter method. Availability: A companion website can be accessed at http://www.ntu. edu.sg/home5/pg02776030/lsbound/. The website contains: (1) the source code of the gene selection algorithm; (2) the complete set of tables and figures regarding the experimental study; (3) proof of the inequality (9).

No abstract

No abstract

In stepped wedge designs (SWD), clusters are randomized to the time period during which new patients will receive the intervention under study in a sequential rollout over time. By the study's end, patients at all clusters receive the intervention, eliminating ethical concerns related to withholding potentially efficacious treatments. This is a practical option in many large-scale public health implementation settings. Little statistical theory for these designs exists for binary outcomes. To address this, we utilized a maximum likelihood approach and developed numerical methods to determine the asymptotic power of the SWD for binary outcomes. We studied how the power of a SWD for detecting risk differences varies as a function of the number of clusters, cluster size, the baseline risk, the intervention effect, the intra-cluster correlation coefficient, and the time effect. We studied the robustness of power to the assumed form of the distribution of the cluster random effects, as well as how power is affected by variable cluster size. % SWD power is sensitive to neither, in contrast to the parallel cluster randomized design which is highly sensitive to variable cluster size. We also found that the approximate weighted least square approach of Hussey and Hughes (2007, Design and analysis of stepped wedge cluster randomized trials. Contemporary Clinical Trials 28, 182-191) for binary outcomes under-estimates the power in some regions of the parameter spaces, and over-estimates it in others. The new method was applied to the design of a large-scale intervention program on post-partum intra-uterine device insertion services for preventing unintended pregnancy in the first 1.5 years following childbirth in Tanzania, where it was found that the previously available method under-estimated the power.