5 papers
Flexible Inference for Winners with Conditional Validity
Soham Bakshi, Lingjun Gao, Zijun Gao +1
Researchers often select top-performing options or winners, based on a data-driven criterion, such as treatments, models, or model features and then report effect estimates for the…
Classification Trees with Valid Inference via the Exponential Mechanism
Soham Bakshi, Snigdha Panigrahi
Decision trees are widely used for non-linear modeling, as they capture interactions between predictors while producing inherently interpretable models. Despite their popularity, p…
From Collapse to Improvement: Statistical Perspectives on the Evolutionary Dynamics of Iterative Training on Contaminated Sources
Soham Bakshi, Sunrit Chakraborty
The problem of model collapse has presented new challenges in iterative training of generative models, where such training with synthetic data leads to an overall degradation of pe…
Inference with Randomized Regression Trees
Soham Bakshi, Yiling Huang, Snigdha Panigrahi +1
Regression trees are a popular machine learning algorithm that fit piecewise constant models by recursively partitioning the predictor space. This paper focuses on statistical infe…
Selective Inference for Time-Varying Moderated Effects
Soham Bakshi, Walter Dempsey, Snigdha Panigrahi
Causal effect moderation investigates how the effect of interventions (or treatments) on outcome variables changes based on observed characteristics of individuals, known as potent…