Research Interests & Experience
My general research interests span neuroscience and engineering. I am particularly motivated by the following questions:
What enables us to conceptualize the world and manipulate it with such abstraction?
Given a goal, how do we select what actions to execute? How was that goal chosen in the first place?
While the primary visual and auditory cortices map the span of frequencies of light and sound, respectively, what would be the analogous parameter space for motor commands and goal representations?
Description of Doctoral Thesis
In a dynamic setting both external environmental stimuli and internally generated actions can yield varying outcomes. Hence, to acquire rewards efficiently an agent needs to track contingencies between stimuli and outcomes as well as actions and outcomes (SO associations and AO associations, respectively). This credit assignment problem is a common challenge in systems relying on artificial intelligence, yet easily solved by biological systems.
Studies in neuroscience have pinpointed the prefrontal cortex to be instrumental in handling the credit assignment problem for controlling goal-oriented behavior. Particularly, lesion studies have shown the need for orbital prefrontal cortex in using SO associations to guide behavior and the need for medial prefrontal cortex in using AO associations to guide behavior. My project characterizes the neuronal representations of these associations underlying the lesion results. Specifically, what mechanisms determine how credit is assigned to stimuli and actions to guide goal-directed behavior?
Understanding these mechanisms may lead to treatments that can prevent people from becoming addicted to drugs of abuse. Some investigators have argued that addiction initially involves administering a drug in a goal-directed manner (we administer the drug because it gives us pleasure) but over time credit assigned to drug-use is ignored and we administer the drug as if on autopilot. Or as a Japanese proverb states, “First the man takes a drink. Then the drink takes a drink. Then the drink takes the man.”
Funding: National Institutes of Health’s Ruth L. Kirschstein National Research Service Award (NRSA) 1F31DA026352
Research Experience
Doctoral thesis on decision-making • University of California, Berkeley • 2005-present
Studying the neuronal mechanisms underlying decision-making – specifically how subjects learn the value of changing environmental cues and their actions
- Collected and analyzed neuronal data from 4 prefrontal brain areas simultaneously
- Trained 6 colleagues to perform neuronal recordings
- Provided technical consultation to peers for their Parkinson’s study
- Writing software tailored to our laboratory to analyze local field potentials in the brain, a large data set with over 30 GB per experiment
- Communicating my work: 1 refereed paper, 4 invited talks, 11 posters (1 invited)
Undergraduate honors thesis on olfaction • University of California, Berkeley • 2004-2005
- Mapped odor perception based on qualitative descriptions into a 4D space using multi-dimension reduction techniques. We found the space to predict similarities across odors well
Awarded honorable mention for student work at the 35th annual meeting of the Measurement Science Conference, Anaheim, CA
- Collaborated with Nanomix, Inc. to study biomimicry on their electronic nose product made of carbon nanotubes. Obtained self-funding through the university’s Guidant Summer Research Program
- Authored 1 refereed paper, 1 invited talk, and 3 posters at national and international conferences
Undergraduate research in microsystems • Berkeley Sensor and Actuator Center, Berkeley, CA • 2002-2003
- Independently built optical tweezers, which used a highly focused laser to physically move micron-sized objects
- Designed a micron-sized lens array that improved detection of cell specimens on a 1 cm2 medical chip by 28 times