Oluwaseun

Oluwaseun

Ph.D. Candidate at Illinois Institute of Technology

Posts by Collection

portfolio

Combinatorial optimization learning for backhauling in B5G/6G wireless networks

The multi-hop multi-commodity flow optimization problem is popular in wireless network optimization, but of more interest in UAV-assisted B5G/6G wireless networks where the objective is to achieve energy-efficient backhauling under certain network constraints. In this project, we develop COLA - a machine learning framework that integrates a combinatorial component into a neural network to achieve energy-efficient backhauling with significantly reduced computational overhead.

Self-renewal machine learning framework for wireless network optimization

We proposed a self-renewal ML (SRML) approach for the optimization of wireless network capacity. The SRML method incrementally improves the throughput maximization of future optimization instances through the design of a data selection algorithm for fine-tuning an application identification model. Our proposed SRML method reduces the computational complexity and achieves a higher solution efficiency.

Bayesian networks and machine learning for COVID-19 severity explanation

We developed a three-stage data-driven approach to distill the hidden information about COVID-19. The first stage employs a Bayesian network structure learning method to identify the causal relationships among COVID-19 symptoms and their intrinsic demographic variables. As a second stage, the output from the Bayesian network structure learning, serves as a useful guide to train an unsupervised machine learning (ML) algorithm that uncovers the similarities in patients’ symptoms through clustering. The final stage then leverages the labels obtained from clustering to train a demographic symptom identification model which predicts a patient’s symptom class and the corresponding demographic probability distribution.

Self-supervised learning approach for accelerating wireless network optimization

In this project, we study the classic wireless multi-commodity network flow problem, where the network-layer routing issue and link-layer scheduling issue are coupled together. We proposed a self-supervised learning paradigm in which the scheduling structures (referring to the independent sets) of historical optimization experiences from a fixed network topology are smartly exploited to facilitate solving new instances with greatly reduced computational overhead.

publications

talks

teaching

Teaching experience 1

Undergraduate course, University 1, Department, 2014

This is a description of a teaching experience. You can use markdown like any other post.

Teaching experience 2

Workshop, University 1, Department, 2015

This is a description of a teaching experience. You can use markdown like any other post.