arXiv subjects

Omar Costilla-Reyes

Publications and source records attributed to Omar Costilla-Reyes.

10 recordsLinked to original sources

arXiv2026

LLMs as Post-hoc Auditors of Physiological Plausibility in Symbolic Regression: A Clinician-Evaluated Case Study

Genetic Programming and its variants, such as grammatical evolution, are widely used in Symbolic Regression to derive mathematical expressions from multivariate data. In addition to predictive accuracy, models are appreciated for their potential to provide interpretability, offering explicit equations that relate input variables to outcomes. However, achieving interpretability and plausibility remains challenging, as evolved models may be complex or scientifically inconsistent. In this study, we explore whether Large Language Models, can assist in improving the explainability of Symbolic Regression models generated by evolutionary computation methods. Building upon our previous work on estimating body fat percentage using grammar-based Genetic Programming , we investigate the use of LLMs as post-processing tools to analyze and rank evolved expressions according to their interpretability and medical plausibility. Four symbolic expressions are analysed by three LLMs over three repeated runs, and the resulting interpretations and rankings are assessed by a panel of three clinicians. Across the three LLMs, comparative model-ranking outputs received more favorable clinician assessments than isolated term-level interpretations. However, the LLMs also produced physiologically and mathematically questionable explanations, indicating that they are better suited to comparative auditing under expert oversight than to autonomous validation.\blfootnote{The present work is an extended version of a paper submitted into a journal.

cs.AI↗

arXiv2025

Predicting Treatment Response in Body Dysmorphic Disorder with Interpretable Machine Learning

Body Dysmorphic Disorder (BDD) is a highly prevalent and frequently underdiagnosed condition characterized by persistent, intrusive preoccupations with perceived defects in physical appearance. In this extended analysis, we employ multiple machine learning approaches to predict treatment outcomes -- specifically treatment response and remission -- with an emphasis on interpretability to ensure clinical relevance and utility. Across the various models investigated, treatment credibility emerged as the most potent predictor, surpassing traditional markers such as baseline symptom severity or comorbid conditions. Notably, while simpler models (e.g., logistic regression and support vector machines) achieved competitive predictive performance, decision tree analyses provided unique insights by revealing clinically interpretable threshold values in credibility scores. These thresholds can serve as practical guideposts for clinicians when tailoring interventions or allocating treatment resources. We further contextualize our findings within the broader literature on BDD, addressing technology-based therapeutics, digital interventions, and the psychosocial determinants of treatment engagement. An extensive array of references situates our results within current research on BDD prevalence, suicidality risks, and digital innovation. Our work underscores the potential of integrating rigorous statistical methodologies with transparent machine learning models. By systematically identifying modifiable predictors -- such as treatment credibility -- we propose a pathway toward more targeted, personalized, and ultimately efficacious interventions for individuals with BDD.

cs.LG↗

arXiv2025

Estimation of total body fat using symbolic regression and evolutionary algorithms

Body fat percentage is an increasingly popular alternative to Body Mass Index to measure overweight and obesity, offering a more accurate representation of body composition. In this work, we evaluate three evolutionary computation techniques, Grammatical Evolution, Context-Free Grammar Genetic Programming, and Dynamic Structured Grammatical Evolution, to derive an interpretable mathematical expression to estimate the percentage of body fat that are also accurate. Our primary objective is to obtain a model that balances accuracy with explainability, making it useful for clinical and health applications. We compare the performance of the three variants on a public anthropometric dataset and compare the results obtained with the QLattice framework. Experimental results show that grammatical evolution techniques can obtain competitive results in performance and interpretability.

cs.NE↗

arXiv2024

MathDSL: A Domain-Specific Language for Concise Mathematical Solutions Via Program Synthesis

We present MathDSL, a Domain-Specific Language (DSL) for mathematical equation solving, which, when deployed in program synthesis models, outperforms state-of-the-art reinforcement-learning-based methods. We also introduce a quantitative metric for measuring the conciseness of a mathematical solution and demonstrate the improvement in the quality of generated solutions compared to other methods. Our system demonstrates that a program synthesis system (DreamCoder) using MathDSL can generate programs that solve linear equations with greater accuracy and conciseness than using reinforcement learning systems. Additionally, we demonstrate that if we use the action spaces of previous reinforcement learning systems as DSLs, MathDSL outperforms the action-space-DSLs. We use DreamCoder to store equation-solving strategies as learned abstractions in its program library and demonstrate that by using MathDSL, these can be converted into human-interpretable solution strategies that could have applications in mathematical education.

cs.LG↗

arXiv2024

Comorbid anxiety predicts lower odds of MDD improvement in a trial of smartphone-delivered interventions

Comorbid anxiety disorders are common among patients with major depressive disorder (MDD), but their impact on outcomes of digital and smartphone-delivered interventions is not well understood. This study is a secondary analysis of a randomized controlled effectiveness trial (n=638) that assessed three smartphone-delivered interventions: Project EVO (a cognitive training app), iPST (a problem-solving therapy app), and Health Tips (an active control). We applied classical machine learning models (logistic regression, support vector machines, decision trees, random forests, and k-nearest-neighbors) to identify baseline predictors of MDD improvement at 4 weeks after trial enrollment. Our analysis produced a decision tree model indicating that a baseline GAD-7 questionnaire score of 11 or higher, a threshold consistent with at least moderate anxiety, strongly predicts lower odds of MDD improvement in this trial. Our exploratory findings suggest that depressed individuals with comorbid anxiety have reduced odds of substantial improvement in the context of smartphone-delivered interventions, as the association was observed across all three intervention groups. Our work highlights a methodology that can identify interpretable clinical thresholds, which, if validated, could predict symptom trajectories and inform treatment selection and intensity.

q-bio.QM↗

arXiv2024

Symbolic Regression with a Learned Concept Library

We present a novel method for symbolic regression (SR), the task of searching for compact programmatic hypotheses that best explain a dataset. The problem is commonly solved using genetic algorithms; we show that we can enhance such methods by inducing a library of abstract textual concepts. Our algorithm, called LaSR, uses zero-shot queries to a large language model (LLM) to discover and evolve concepts occurring in known high-performing hypotheses. We discover new hypotheses using a mix of standard evolutionary steps and LLM-guided steps (obtained through zero-shot LLM queries) conditioned on discovered concepts. Once discovered, hypotheses are used in a new round of concept abstraction and evolution. We validate LaSR on the Feynman equations, a popular SR benchmark, as well as a set of synthetic tasks. On these benchmarks, LaSR substantially outperforms a variety of state-of-the-art SR approaches based on deep learning and evolutionary algorithms. Moreover, we show that LaSR can be used to discover a novel and powerful scaling law for LLMs.

cs.LG↗

arXiv2023

Individual Behavioral Insights in Schizophrenia: A Network Analysis and Mobile Sensing Approach

Digital phenotyping in mental health often consists of collecting behavioral and experience-based information through sensory and self-reported data from devices such as smartphones. Such rich and comprehensive data could be used to develop insights into the relationships between daily behavior and a range of mental health conditions. However, current analytical approaches have shown limited application due to these datasets being both high dimensional and multimodal in nature. This study demonstrates the first use of a principled method which consolidates the complexities of subjective self-reported data (Ecological Momentary Assessments - EMAs) with concurrent sensor-based data. In this study the CrossCheck dataset is used to analyse data from 50 participants diagnosed with schizophrenia. Network Analysis is applied to EMAs at an individual (n-of-1) level while sensor data is used to identify periods of various behavioral context. Networks generated during periods of certain behavioral contexts, such as variations in the daily number of locations visited, were found to significantly differ from baseline networks and networks generated from randomly sampled periods of time. The framework presented here lays a foundation to reveal behavioural contexts and the concurrent impact of self-reporting at an n-of-1 level. These insights are valuable in the management of serious mental illnesses such as schizophrenia.

cs.SI↗

arXiv2023

Counterfactual Explanations and Predictive Models to Enhance Clinical Decision-Making in Schizophrenia using Digital Phenotyping

Clinical practice in psychiatry is burdened with the increased demand for healthcare services and the scarce resources available. New paradigms of health data powered with machine learning techniques could open the possibility to improve clinical workflow in critical stages of clinical assessment and treatment in psychiatry. In this work, we propose a machine learning system capable of predicting, detecting, and explaining individual changes in symptoms of patients with Schizophrenia by using behavioral digital phenotyping data. We forecast symptoms of patients with an error rate below 10%. The system detects decreases in symptoms using changepoint algorithms and uses counterfactual explanations as a recourse in a simulated continuous monitoring scenario in healthcare. Overall, this study offers valuable insights into the performance and potential of counterfactual explanations, predictive models, and change-point detection within a simulated clinical workflow. These findings lay the foundation for further research to explore additional facets of the workflow, aiming to enhance its effectiveness and applicability in real-world healthcare settings. By leveraging these components, the goal is to develop an actionable, interpretable, and trustworthy integrative decision support system that combines real-time clinical assessments with sensor-based inputs.

cs.AI↗

arXiv2022

LEMMA: Bootstrapping High-Level Mathematical Reasoning with Learned Symbolic Abstractions

Humans tame the complexity of mathematical reasoning by developing hierarchies of abstractions. With proper abstractions, solutions to hard problems can be expressed concisely, thus making them more likely to be found. In this paper, we propose Learning Mathematical Abstractions (LEMMA): an algorithm that implements this idea for reinforcement learning agents in mathematical domains. LEMMA augments Expert Iteration with an abstraction step, where solutions found so far are revisited and rewritten in terms of new higher-level actions, which then become available to solve new problems. We evaluate LEMMA on two mathematical reasoning tasks--equation solving and fraction simplification--in a step-by-step fashion. In these two domains, LEMMA improves the ability of an existing agent, both solving more problems and generalizing more effectively to harder problems than those seen during training.

cs.AI↗

arXiv2022

Neurosymbolic Programming for Science

Neurosymbolic Programming (NP) techniques have the potential to accelerate scientific discovery. These models combine neural and symbolic components to learn complex patterns and representations from data, using high-level concepts or known constraints. NP techniques can interface with symbolic domain knowledge from scientists, such as prior knowledge and experimental context, to produce interpretable outputs. We identify opportunities and challenges between current NP models and scientific workflows, with real-world examples from behavior analysis in science: to enable the use of NP broadly for workflows across the natural and social sciences.

cs.AI↗