About Research Publications Current Work Students Teaching Awards Contact
Assistant Professor · Data Science

Bridging Rigorous Mathematics and Applied Machine Learning

I am an Assistant Professor of Data Science in the Department of Mathematics & Statistics at San José State University. My research spans two interconnected areas: the nonlinear analysis of PDEs and chemotaxis models, and applied machine learning — including deep learning, NLP, and computer vision — with applications in biomedical imaging, climate science, and blockchain systems.

Before joining SJSU, I was a Data Science Postdoctoral Fellow at The Data Institute, University of San Francisco (2019–2021), collaborating with Harvard Center for Systems Biology and MedStar Georgetown University Hospital on GAN-based microscopy and clinical NLP pipelines. I hold a Ph.D. in Mathematics from Auburn University, advised by Prof. Wenxian Shen.

⚗ Nonlinear PDEs 🧠 Deep Learning 💬 NLP 🔬 Computer Vision 📊 Statistics 🌊 Chemotaxis Models 🔥 Fire Weather Risk ⛓ Blockchain & AI
View Publications Contact Me
Tahir Bachar Issa

Tahir Bachar Issa

Assistant Professor of Data Science
Dept. of Mathematics & Statistics
San José State University · San José, CA
🔢

Applied Mathematics

Nonlinear PDE analysis, chemotaxis and reaction-diffusion systems, stability theory, and mathematical biology.

🤖

Machine Learning

Deep learning, GANs, LLM fine-tuning, activation function research, and ensemble methods for real-world deployment.

🔬

Biomedical Imaging

Super-resolution microscopy via GANs, endothelial mitochondria segmentation, and clinical NLP for radiology.

📡

Data Science

NLP, computer vision, extreme weather ML, blockchain analytics, and production model deployment.

Recent Work

Research Highlights

⚡

SAW: High-Performing Activation Functions for Deep Learning

Novel activation functions outperforming state-of-the-art for training deep learning models and LLMs. Submitted to CIKM 2026.

S. Mohanty, T. B. Issa, M. Masum, D. Uminsky · 2026
Under Review
🧮

Uniqueness & Nonlinear Stability in Full Chemotaxis Models

Rigorous stability analysis for entire solutions in full chemotaxis models with logistic source on bounded heterogeneous environments.

T. B. Issa · Proc. AMS Ser. B, 13 (2026), pp. 44–58
DOI ↗
🔥

ML-Based Prediction of Extreme Fire Weather Risk

Ensemble machine learning to characterize extreme fire weather risk characteristics across the contiguous United States.

H. R. Seethagari, I. Diallo, T. B. Issa · SJSU RSD21, 2026
🏥

Label-Free Endothelial Cell Mitochondria Segmentation

Holotomographic microscopy combined with UNet for label-free visualization and segmentation of mitochondrial networks.

R. Michael, T. Modirzadeh, T. B. Issa, P. Jurney · Chem. Biomed. Imaging, 2025
DOI ↗
🦠

Two-Strain SIS Epidemic Model in Patchy Environments

Competition-exclusion and coexistence dynamics in a two-strain epidemic model, analyzing long-run persistence across heterogeneous patches.

J. T. Doumaté, T. B. Issa, R. B. Salako · DCDS-B, 29(7), 2024
DOI ↗
🔭

Super-Resolution Fluorescence Microscopy via GANs

Real-world benchmarks for super-resolving fluorescence microscope imagery using Generative Adversarial Networks. Harvard collaboration.

J. Cooper, T. B. Issa, C. Vinegoni, R. Weissleder · IEEE CAI 2024
IEEE ↗
Full List

Publications & Preprints

2026
1
SAW: High Performing Activation Functions for Training Deep Learning Models
S. Mohanty, T. B. Issa, M. Masum, D. Uminsky
Submitted — 35th International ACM Conference on Knowledge and Information Management (CIKM 2026)
Under Review
2
Uniqueness and nonlinear stability of entire solutions in full chemotaxis models with logistic source on bounded heterogeneous environments
Tahir B. Issa
Proc. Amer. Math. Soc. Ser. B, 13 (2026), pp. 44–58
DOI AMS
3
Colorectal Cancer Detection using Histopathological Images
N. S. Chilumukuru, S. Sriramulu, T. B. Issa, Y. Ezunkpe
2026
4
New State-of-the-Art Activation Functions for Training LLMs
S. Mohanty, T. B. Issa
SJSU RSD21 Poster, 2026
5
Machine Learning Based Prediction of Extreme Fire Weather Risk over the Contiguous United States
H. R. Seethagari, I. Diallo, T. B. Issa
SJSU RSD21 Poster, 2026
2025
6
Label-Free Visualization and Segmentation of Endothelial Cell Mitochondria Using Holotomographic Microscopy and UNet
R. Michael, T. Modirzadeh, T. B. Issa, P. Jurney
Chemistry & Biomedical Imaging, 2025
DOI
2024
7
Providing Real-World Benchmarks for Super-Resolving Fluorescence Microscope Imagery Using Generative Adversarial Networks
J. Cooper, T. B. Issa, C. Vinegoni, R. Weissleder
2024 IEEE Conference on Artificial Intelligence (CAI), Singapore, pp. 1154–1161
IEEE
8
Competition-exclusion and coexistence in a two-strain SIS epidemic model in patchy environments
J. T. Doumaté, T. B. Issa, R. B. Salako
Discrete and Continuous Dynamical Systems – Series B, 29(7): 3058–3096, 2024
DOI
2023
9
Using Machine Learning to Model Potential Users with Health Risk Concerns Regarding Microchip Implants
S. Shafeie, M. A. Mohamed, T. B. Issa, B. M. Chaudhry
Artificial Intelligence in HCI, HCIII 2023, Copenhagen, Proceedings Part II
DOI
10
Deep learning-based numerical solutions and their theoretical stability for parabolic-parabolic chemotaxis models with nonlocal logistic sources
T. B. Issa, Y. Ezunkpe
1st International Online Conference on Mathematics and Applications, MDPI, 2023
DOI
2022
11
Protocol Selection of Advanced Imaging Exams using Multi-steps Deep Learning Models
C. Cronister, T. B. Issa, D. Uminsky, R. W. Filice
Open Science Index 16, Vol 2, p. 346, 2022
12
Protocol Selection of Advanced Imaging Exams using Multi-steps Deep Learning Models
C. Cronister, T. B. Issa, D. Uminsky, R. W. Filice
15th IEEE EMBS Regional Conference (CISP-BMEI 2022)
2021
13
Traveling wave solutions for two species competitive chemotaxis systems
T. B. Issa, R. B. Salako, W. Shen
Nonlinear Analysis, 212, 2021
DOI
14
Toward Automatic Mammography Auditing via Universal Language Model Fine Tuning
T. B. Issa, D. Uminsky, A. Shaw, R. Makipour, R. W. Filice
2021 IEEE 22nd International Conference on Information Reuse and Integration (IRI), pp. 215–222
IEEE
15
Predicting microcystin occurrence in freshwater lakes and reservoirs: assessing environmental variables
R. P. Buley, H. E. Correia, A. Abebe, T. B. Issa, A. E. Wilson
Inland Waters, 11(3), 2021
2020
16
Video-rate acquisition fluorescence microscopy via generative adversarial networks
T. B. Issa, C. Vinegoni, A. Shaw, P. F. Feruglio, R. Weissleder, D. Uminsky
IEEE 20th International Conference on Bioinformatics and Bioengineering (BIBE 2020), pp. 569–576
IEEE
17
A hybrid data analytics approach for high-performance concrete compressive strength prediction
S. Simsek, M. Gumus, M. Khalafalla, T. B. Issa
Journal of Business Analytics, 2020, pp. 1–11
18
Pointwise persistence in full chemotaxis models with logistic source on bounded heterogeneous environments
T. B. Issa, W. Shen
Journal of Mathematical Analysis and Applications, 490(1), 124204, 2020
2018 – 2017
19
Uniqueness and stability of coexistence states in two species models with/without chemotaxis on bounded heterogeneous environments
T. B. Issa, W. Shen
Journal of Dynamics and Differential Equations, 2018
DOI
20
Persistence, coexistence and extinction in two species chemotaxis models on bounded heterogeneous environments
T. B. Issa, W. Shen
Journal of Dynamics and Differential Equations, 2018
DOI
21
Dynamics in chemotaxis models of parabolic-elliptic type on bounded domain with time and space dependent logistic sources
T. B. Issa, W. Shen
SIAM Journal on Applied Dynamical Systems, 16(2), 926–973, 2017
22
Asymptotic dynamics in a two-species chemotaxis model with non-local terms
T. B. Issa, R. B. Salako
Discrete and Continuous Dynamical Systems – Series B, 22(5), 3839–3874, 2017
In Preparation
P1
Dynamics of parabolic-ODE chemotaxis models on unbounded homogeneous and heterogeneous environments
T. B. Issa
P2
Colorectal Cancer Detection using Histopathological Images
N. S. Chilumukuru, S. Sriramulu, T. B. Issa, Y. Ezunkpe
P3
New State-of-the-Art Activation Functions for Training LLMs
S. Mohanty, T. B. Issa
P4
Machine Learning Based Prediction of Extreme Fire Weather Risk over the Contiguous United States
H. R. Seethagari, I. Diallo, T. B. Issa
Active Projects

Current Work

2026 – Present
SAW Activation Functions for Deep Learning & LLMs
with Siddhant Mohanty · SJSU · Submitted CIKM 2026
Developing mathematically grounded activation functions that improve training dynamics and generalization in deep learning models and large language models.
2026 – Present
ML-Based Prediction of Extreme Fire Weather Risk
with H. R. Seethagari, I. Diallo · VPRI Teaming Award, Co-PI ($25,000)
Ensemble machine learning methods for characterizing extreme fire weather risk across the contiguous United States.
Fall 2024 – Present
Anti-Censorship Resistance in the Ethereum Blockchain
Ethereum Foundation Grant NK0822-01, Co-PI ($50,000) · CAMCOS, SJSU
Spring 2024 – Present
NVIDIA Deep Learning Institute University Ambassador
San José State University
Bringing GPU computing and deep learning curricula to SJSU through NVIDIA's DLI program.
Mentoring

Student Supervision

At SJSU I mentor graduate and undergraduate students in data science research. During my postdoc at USF (2019–2021), 100% of mentored students secured industry data science positions upon graduation.

Current Students · 2026

Siddhant Mohanty
Graduate Student · SJSU, 2026
New state-of-the-art activation functions for training LLMs — SJSU RSD21 & CIKM 2026
Harshavardhan Reddy Seethagari & Ismaila Diallo
Graduate Students · SJSU, 2026
ML-based prediction of extreme fire weather risk over the contiguous United States — SJSU RSD21

Completed Students

Chun Tam
MSDS · SJSU, 2024–2025
Deep learning for quantification of endothelial cell mitochondrial network structure and dynamics
CAMCOS Team (Applied Materials)
Undergraduate/Graduate · SJSU, Spring 2025
Industry collaboration with Applied Materials
Xiang Yao
MS Statistics · SJSU, Spring & Fall 2024
Generalized state-of-the-art activation functions for deep learning models training
Quoc Luong Huynh
BS Statistics · SJSU, Summer & Fall 2024
Building a state-of-the-art AI-powered Email Replying Assistant App [Demo]
John Cooper
MS Statistics · SJSU, 2023 (now Data Scientist)
A perceptual lens for super-resolving video-rate acquisition microscope imagery
Steven Rogalsky
MS Statistics · SJSU, Fall 2023 – Spring 2024
Audio classification using deep learning
Dat Le & Nicholas Sobrepena
Undergraduate · CAMCOS, SJSU, July 2023
Multidimensional Ethereum Virtual Machine Transaction Modeling [Link]
C. Cronister
Data Institute · University of San Francisco
Protocol selection of advanced imaging exams using multi-steps deep learning models
Background

Education & Training

2019 – 2021
Postdoctoral Fellow, Data Science
The Data Institute, University of San Francisco, CA
2013 – 2019
Ph.D., Mathematics
Auburn University, Auburn, AL
Dissertation: Dynamics of Chemotaxis Models in Heterogeneous Environments · Advisor: Prof. Wenxian Shen
2016 – 2018
M.S., Probability and Statistics
Auburn University, Auburn, AL · Advisor: Prof. Ashe Abebe
2012
Postgraduate Diploma, Mathematics
International Centre for Theoretical Physics (ICTP), Trieste, Italy
Thesis: A Recent Variational Approach to Semilinear Wave Equations · Advisor: Prof. Giovanni Bellettini
2011
M.Sc., Pure and Applied Mathematics
African University of Science and Technology, Nigeria
2010
M.Sc., Applied Mathematics and Computer Science
Gaston Berger University, Senegal
2009
B.Sc., Applied Mathematics and Computer Science
Gaston Berger University, Senegal
Classroom

Teaching Experience

San José State University (2021 – Present)

SemesterCourse
Spring 2026Mathematical Data Visualization
Spring 2026Introduction to Python Programming and SQL
Fall 2025Statistical and Machine Learning Classification
Fall 2024Machine Learning for Classification
Fall 2024Calculus III
Spring 2024Introduction to Python Programming and SQL
Fall 2023Machine Learning for Classification
Fall 2023Probability and Statistics Theory
Spring 2023Fundamentals of Data Science
Fall 2022R Programming for Data Analysis
Fall 2022Applied Probability and Statistics
Fall 2021 / Spring 2022Probability and Statistics Theory & Applied Probability and Statistics

Auburn University — Graduate Teaching Assistant (2013 – 2019)

PeriodCourse
Summer 2019Statistics for Engineers and Scientists
Spring 2018, 2019Statistics for Biological and Health Sciences
Fall 2018Probability and Statistics I
Multiple termsDifferential Equations, Calculus II & III, Linear Algebra
Multiple termsPre-calculus (Math 1000, 1120, 1150)
Recognition

Awards & Grants

Toolkit

Skills & Service

Programming & Tools

PythonPyTorchFastAI RSQLSpark PySparkMATLABCC++

Languages

EnglishFrench Chadian ArabicArabic

Journal Reviewing

Nonlinear Analysis (Elsevier) Nonlinearity (IOP/LMS) Chaos (AIP) Annals of Operations Research Electronic Commerce Research

Professional Affiliations

SIAM (2018–present) American Mathematical Society (2013–present) NVIDIA DLI Ambassador
Get in Touch

Contact Info

I welcome inquiries from prospective students, collaborators, and colleagues.

📧 tahirbachar.issa@sjsu.edu
📞 408-924-5146
📍 One Washington Square, San José, CA 95192-0103
🏛 Dept. of Mathematics & Statistics, SJSU
🎓 Google Scholar 🏛 Dept. Website
Location

San José State University

Tahir Bachar Issa
Assistant Professor of Data Science
Department of Mathematics & Statistics
San José State University
One Washington Square
San José, CA 95192-0103