profile image

Taha Belkhouja

Ph.D. in Computer Science

(Last Update: February 2026)

Contact Me

About Me

I am an AI Applied Scientist with expertise in developing and deploying enterprise-scale Generative AI systems. I have hands-on experience in architecting AI-based solutions and multi-AI-agent systems, with a proven track record of translating research innovations into productionized solutions.

I obtained my Ph.D. in Computer Science at Washington State University where I was advised by Professor Jana Doppa. My general research interests are in the area of robust and trustworthy machine learning. My PhD research focused on developing efficient algorithms and theory to improve reliability and safety of deep learning algorithms for diverse problem settings and data domains.

Projects

Tech lead in AIRefinery

I am a Technical Lead in the design and deployment of AI Refinery™ agentic AI systems that translate advances in generative modeling and reasoning into production-ready platforms, enabling organizations to transform raw AI capabilities into reliable business solutions. I contribute to building and evaluating scalable multi-agent workflows that support complex enterprise decision-making across diverse application domains. This work is complemented by the development of an SDK that allows organizations to build, customize, and securely deploy agentic AI applications across cloud and on-premise environments.

Robustness and reliability for deep learning models

My research develops deep learning algorithms for the safe deployment of machine learning models operating under real-world perturbations such as noise, temporal shifts, and distributional distortions. I design robust training objectives and similarity measures that improve model reliability when standard distance metrics fail under natural or adversarial transformations of the input space. I also investigate methods for detecting out-of-distribution inputs to prevent overconfident predictions in safety-critical environments. I have also worked on theoretically grounded uncertainty quantification techniques, including conformal prediction, to enable calibrated and trustworthy decision-making in human-AI systems.
Highlighted papers:
Adversarial Framework for Time-Series
Out-of-distribution Detection in Time-series Domain
Uncertainty Quantification of Deep Classifiers

Select Publications

Adaptive Data Augmentation with Multi-armed Bandit: Sample-Efficient Embedding Calibration for Implicit Pattern Recognition M. Tang, Y. Yu, A. Ding, M. B. Pouyan, T. Belkhouja, Y. Bao The IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2026

Conformal Prediction for Class-wise Coverage via Augmented Label Rank Calibration Y. Shi, S. Ghosh, T. Belkhouja, J. Doppa, Y. Yan Conference on Neural Information Processing Systems (NeurIPS), 2024

Improving Uncertainty Quantification of Deep Classifiers via Neighborhood Conformal Prediction T. Belkhouja*, S. Ghosh*, Y. Yan, J. Doppa AAAI Conference on Artificial Intelligence, 2023 · (* equal contribution)

Probabilistically Robust Conformal Prediction S. Ghosh, Y. Shi, T. Belkhouja, Y. Yan, J. Doppa, B. Jones 39th Uncertainty in Artificial Intelligence (UAI), 2023

Energy-Efficient Missing Data Recovery in Wearable Devices: A Novel Search-based Approach T. Belkhouja*, D. Hussein*, G. Bhat, J. Doppa ACM/IEEE International Symposium on Low Power Electronics and Design (ISLPED), 2023 · (* equal contribution)

Out-of-Distribution Detection in Time-Series Domain: A Novel Seasonal Ratio Scoring Approach T. Belkhouja, Y. Yan, J. Doppa ACM Transactions on Intelligent Systems and Technology (TIST), 2023

Adversarial Framework with Certified Robustness for Time-Series Data via Statistical Features T. Belkhouja, J. Doppa International Joint Conference on Artificial Intelligence (IJCAI), 2023

Dynamic Time Warping based Adversarial Framework for Time-Series Domain T. Belkhouja, Y. Yan, J. Doppa IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 2022

Training Robust Deep Models for Time-Series Domain: Novel Algorithms and Theoretical Analysis T. Belkhouja, Y. Yan, J. Doppa AAAI Conference on Artificial Intelligence, 2022

Reliable Machine Learning for Wearable Activity Monitoring: Novel Algorithms and Theoretical Guarantees T. Belkhouja*, D. Hussein*, G. Bhat, J. Doppa 41st International Conference on Computer-Aided Design (ICCAD), 2022 · (* equal contribution)

Adversarial Framework with Certified Robustness for Time-Series Data via Statistical Features T. Belkhouja, J. Doppa Journal of Artificial Intelligence Research (JAIR), 2022

Analyzing Deep Learning for Time-Series Data through Adversarial Lens in Mobile and IoT Applications T. Belkhouja, J. Doppa IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems (TCAD), 2020

Service

Professional Events Organization

Tutorial on Advances in Robust Time-Series ML - AAAI 2024. Lead organizer

Program Committee Member

International Conference of Machine Learning - (ICML) 2024

AAAI Conference on Artificial Intelligence - (AAAI) 2024

AAAI Conference on Artificial Intelligence - AAAI Student Program 2024

AAAI Conference on Artificial Intelligence - AAAI Safe and Robust AI Track 2024

Conference on Neural Information Processing Systems - (NeurIPS) 2023

International Conference of Machine Learning - (ICML) 2023

International Conference on Artificial Intelligence and Statistics - (AISTATS) 2023

AAAI Conference on Artificial Intelligence - (AAAI) 2023

Teaching

Teaching Assistant, Washington State University

CptS 315 - Introduction to Data Mining Spring 2020 - Spring 2021

CptS 570 - Machine Learning Fall 2020

CptS 223 - Advanced Data Structures in C++ Fall 2020

CptS 451 - Introduction to Database Systems Spring 2020

CptS 440/540 - Introduction to Data Mining Fall 2019

Teaching Assistant, University of Idaho

ECE 241 - Digital Logic Circuit Lab 2017-2019

ECE 311 - Microelectronics I Lab Spring 2019