서브메뉴
검색
Inferring Latent Factors and States Underlying Behavior and Neural Dynamics
Inferring Latent Factors and States Underlying Behavior and Neural Dynamics
Detailed Information
- 자료유형
- 학위논문 서양
- 최종처리일시
- 20250211151434
- ISBN
- 9798382807195
- DDC
- 621.3
- 저자명
- Jha, Aditi.
- 서명/저자
- Inferring Latent Factors and States Underlying Behavior and Neural Dynamics
- 발행사항
- [Sl] : Princeton University, 2024
- 발행사항
- Ann Arbor : ProQuest Dissertations & Theses, 2024
- 형태사항
- 225 p
- 주기사항
- Source: Dissertations Abstracts International, Volume: 85-12, Section: B.
- 주기사항
- Advisor: Pillow, Jonathan W.
- 학위논문주기
- Thesis (Ph.D.)--Princeton University, 2024.
- 초록/해제
- 요약Rapid advancements in experimental neuroscience are now generating a wealth of high-resolution neural and behavioral datasets. This opens new avenues for understanding brain computations and observed behaviors. Neural and behavioral datasets are complex, often containing several dimensions. However, studies consistently show that a small number of underlying factors can explain much of the observed complexity and offer interpretable descriptions. Nonetheless, accurately extracting these latent factors poses several challenges, particularly given the limited samples in these high-dimensional datasets. This thesis addresses these challenges by presenting several novel statistical approaches that are data-efficient and are tailored for neuroscientific datasets.In the first half, we present methods to uncover and interpret the low-dimensional representations that underlie neural activity and behavior during different perceptual tasks. We also introduce a model class designed to extract the underlying dynamics of different neural populations engaged in sensory tasks. Notably, this framework enables testing the causal involvement of various neural circuits in observed behavior. The second half of this thesis focuses on animal behavior: given the challenge of data collection in neuroscience, we propose an approach aimed at accelerating the inference of internal states that describe observed animal behavior during decision-making. Finally, we develop a novel formulation to understand complex animal behavior from the perspective of an animal's goals and actions, using inverse reinforcement learning.Overall, this thesis highlights the efficacy of data-efficient methods and approaches with the tailored inductive biases towards obtaining a nuanced understanding of neural computations. It advocates for novel paradigms in animal behavior modeling, contributes to the expanding literature on dynamic models of behavior, and adds to recent endeavors in disentangling the roles of different neural populations during a task.
- 일반주제명
- Electrical engineering
- 일반주제명
- Neurosciences
- 일반주제명
- Computer engineering
- 키워드
- Neural dynamics
- 기타저자
- Princeton University Electrical and Computer Engineering
- 기본자료저록
- Dissertations Abstracts International. 85-12B.
- 전자적 위치 및 접속
- 로그인 후 원문을 볼 수 있습니다.
MARC
008250123s2024 us c eng d■001000017161707
■00520250211151434
■006m o d
■007cr#unu||||||||
■020 ▼a9798382807195
■035 ▼a(MiAaPQ)AAI31295466
■040 ▼aMiAaPQ▼cMiAaPQ
■0820 ▼a621.3
■1001 ▼aJha, Aditi.
■24510▼aInferring Latent Factors and States Underlying Behavior and Neural Dynamics
■260 ▼a[Sl]▼bPrinceton University▼c2024
■260 1▼aAnn Arbor▼bProQuest Dissertations & Theses▼c2024
■300 ▼a225 p
■500 ▼aSource: Dissertations Abstracts International, Volume: 85-12, Section: B.
■500 ▼aAdvisor: Pillow, Jonathan W.
■5021 ▼aThesis (Ph.D.)--Princeton University, 2024.
■520 ▼aRapid advancements in experimental neuroscience are now generating a wealth of high-resolution neural and behavioral datasets. This opens new avenues for understanding brain computations and observed behaviors. Neural and behavioral datasets are complex, often containing several dimensions. However, studies consistently show that a small number of underlying factors can explain much of the observed complexity and offer interpretable descriptions. Nonetheless, accurately extracting these latent factors poses several challenges, particularly given the limited samples in these high-dimensional datasets. This thesis addresses these challenges by presenting several novel statistical approaches that are data-efficient and are tailored for neuroscientific datasets.In the first half, we present methods to uncover and interpret the low-dimensional representations that underlie neural activity and behavior during different perceptual tasks. We also introduce a model class designed to extract the underlying dynamics of different neural populations engaged in sensory tasks. Notably, this framework enables testing the causal involvement of various neural circuits in observed behavior. The second half of this thesis focuses on animal behavior: given the challenge of data collection in neuroscience, we propose an approach aimed at accelerating the inference of internal states that describe observed animal behavior during decision-making. Finally, we develop a novel formulation to understand complex animal behavior from the perspective of an animal's goals and actions, using inverse reinforcement learning.Overall, this thesis highlights the efficacy of data-efficient methods and approaches with the tailored inductive biases towards obtaining a nuanced understanding of neural computations. It advocates for novel paradigms in animal behavior modeling, contributes to the expanding literature on dynamic models of behavior, and adds to recent endeavors in disentangling the roles of different neural populations during a task.
■590 ▼aSchool code: 0181.
■650 4▼aElectrical engineering
■650 4▼aNeurosciences
■650 4▼aComputer engineering
■653 ▼aBehavioral datasets
■653 ▼aNeural dynamics
■653 ▼aStatistical models
■653 ▼aBrain computation
■690 ▼a0544
■690 ▼a0317
■690 ▼a0464
■71020▼aPrinceton University▼bElectrical and Computer Engineering.
■7730 ▼tDissertations Abstracts International▼g85-12B.
■790 ▼a0181
■791 ▼aPh.D.
■792 ▼a2024
■793 ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17161707▼nKERIS▼z이 자료의 원문은 한국교육학술정보원에서 제공합니다.
Preview
Export
ChatGPT Discussion
AI Recommended Related Books
פרט מידע
- הזמנה
- לא קיים
- התיקיה שלי
- צפה הראשון בקשה
- Non-Book Loan Application
- Nighttime Book Loan Application
Available after logging in.


