سيف حسام رحيم

This paper presents a Bayesian framework for estimating individualized treatment effects (ITE) in high-dimensional observational data. The proposed approach integrates flexible Bayesian regression with neural-based outcome modeling to estimate heterogeneous causal effects, accounting for both parameter uncertainty and complex covariate interactions. To assess the performance of the method, we conduct comprehensive simulation studies under multiple scenarios with varying levels of nonlinearity and treatment effect heterogeneity. Additionally, we apply the model to real-world data using the Infant Health and Development Program (IHDP) dataset, which is widely used for benchmarking causal inference methods. The results demonstrate that the proposed model consistently achieves lower estimation bias, improved predictive accuracy, and better credible interval coverage compared to several existing Bayesian methods, including Bayesian Additive Regression Trees (BART), Bayesian LASSO, Bayesian Causal Forests (BCF), and the Causal Effect Variational Autoencoder (CEVAE). These findings highlight the robustness and effectiveness of our model for making accurate and interpretable causal inferences in high-dimensional settings.

High-dimensional gene expression data pose major challenges for statistical modeling due to the large number of predictors, strong correlations, and the presence of nonlinear regulatory structures. This study proposes a hybrid Bayesian framework that combines Bayesian Additive Regression Trees (BART) with the Reciprocal LASSO prior to achieve flexible nonlinear modeling and structured sparsity within a unified model. Theoretical development integrates aggressive shrinkage for variable selection with a sum-of-trees architecture that captures complex gene–gene interactions. A comprehensive simulation study across multiple dimensional and correlation settings demonstrates that the proposed BART–RL model consistently achieves lower prediction error and higher true positive rates compared with classical LASSO, elastic net, BART, and Bayesian reciprocal LASSO. Application to a real gene expression dataset further confirms the advantages of the hybrid approach, yielding improved predictive performance and identifying biologically meaningful genes supported by functional annotations. These results highlight the utility of combining nonlinear Bayesian tree ensembles with adaptive shrinkage priors for high-dimensional genomic modeling.

.

The authors propose robust group-lasso for sliced inverse regression in this paper (robust group lasso-SIR). This proposed approach will deal with the association problem that occurs between predictor variables. Simulation is used to evaluate the achievement of the proposed approach relative to a sliced inverse regression lasso (lasso-SIR) and a robust lasso-SIR group (robust group lasso-SIR). The findings show that, based on the Mean Square Errors (MSE) test, the stable group lasso-SIR approach performs well relative to different methods.

Restricted data plays a significant role in describing economic, social, medical, with other phenomena. Most of the time the Restricted point on zero, so the appropriate regression model for this type of data is a Tobit regression model. When the number of independent variables is too large, the process of their interpretation is very complex. To get around this problem, it is possible to use Variable Selections. In the current paper, we will use the adaptive Lasso through the Bayesian method. Also, the Bayesians Lasso method has many advantages that provide accuracy in the results, especially in the selection of Variable Selections. To compare our proposal, we will use the number of infections with Covid-19 for a group of families through a field survey in Al-Qadisiyah Governorate and identify the effective factors.

.

.