You are here

Faculty Spotlight: Parinaz Naghizadeh

As autonomous technology advances, critics are quick to point out the downsides. It could make highway traffic worse, render hacking and cyberattacks easier, or make privacy even more unreliable. So, scientists want to know: what is the cause and effect of the decisions made as humans rely more upon machines?
 
Parinaz Naghizadeh is a new assistant professor in both Integrated Systems and Electrical and Computer Engineering at The Ohio State University. She’s looking at this question mathematically, to help the world learn more about how machines and humans interact.
 
“Their decisions affect one another. The way they learn and evolve affects one another,” Naghizadeh said. “I look at how this type of interdependent learning and decision making happens. More precisely, I build mathematical models that analyze this critically.”
 
Naghizadeh is also fascinated by the process of how machines learn to accomplish tasks on their own.
 
“There are really fun things machines do when you ask them to learn by themselves,” she said. “They tend to find ways to do things that, in hindsight, look very reasonable to us, but we may not have thought about putting it in the order they do.” 

The field studying these processes is called “reinforcement learning,” and she explains why it could be important for students at Ohio State to think of as a potential career.
 
“There is a lot of data available, and we are developing more and more autonomous systems. We are still figuring out how to make all these systems work and how they affect humans; figuring out good ways to use the data,” Naghizadeh said. “Reinforcement learning is this emerging field that says I can almost leave out this robot or this autonomous agent with a task, and they can learn how to handle this task through trial and error. I think that’s a very good concept for our students to know and use in designing autonomous systems. We don’t want to just program these systems to automate tasks with a controller or with static decision making. We want them to be more interactive. I think interactive learning or reinforcement learning is a very exciting field because of that.”
 
As a young student, Naghizadeh excelled when working with numbers and today studies the gray area between engineering and economics.
 
“Math was my favorite subject in high school, and then I went into electrical engineering,” she said. “It is such a broad field. You can be building communication systems and nanotech devices, or you can be taking a more theoretical approach to understanding engineered systems. That was the path I took as an undergrad and it all led to focusing on mathematical models of these systems in my research.”
 
Naghizadeh came to Ohio State to advance her work in decision sciences and control, studying complex systems and networks. She is teaching optimization and reinforcement learning.
 
“There are really strong people here in my areas of research, both in ECE and ISE,” she said. “It’s a good place to be.”