Anubhab Ganguly

Experience

• 

  International Youth Neuroscience Association

  Mar 2018 - now

  Member
• 

  Neuromatch Academy

  May 2022 - Sept 2022

  Student

  • It is a 3 week, online, worldwide, immersive summer school in Computational Neuroscience/ Machine Learning taught by world leaders in the field. • The cohort is focused on physics, mathematics and coding required to understand and develop computational models in neuroscience. We are taught model types, model fitting (Linear Regression, Polynomial Regression, MSE, MLE, Least Squares Optimization), Generalized Linear Models, dimensionality reduction, dynamical systems and differential equations. The hands-on tutorials are done in Google Colab. Show less

• 

  Massachusetts General Hospital, Martinos Center for Biomedical Imaging

  Jan 2023 - Mar 2023

  Harvard Aging Brain Study(HABS) Shadow Program Attendee

  • Learned about observational and clinical trials related to Alzheimer’s disease for 6 weeks. • Learned about the various components of studies on Alzheimer’s disease, particularly neuropsychological testing, imaging and recruitment. • Gained first-hand experience administering blood pressure and neuropsychological testing: Trail Making Test A and B, Mini Mental State Exam. • Received mentorship from various Faculty and Staff within the Harvard Medical School. Show less

• 

  Independent Project

  May 2023 - Aug 2023

  Student

  Project- Attractor Model of short-term memory• In this project, I have built a simple, population based, dynamical systems model of a short-term memory task.• Short-term memory is persistent, stimulus-dependent neuronal activity observed in some brain areas requiring the temporary maintenance of information.• I have simulated a dynamical system consisting of two excitatory neurons and one inhibitory neuron connected to both the excitatory neurons. I used firing rate equations and sigmoidal non- linearity as transfer function.• I used Wilson-Cowan equations to model the activity of the neurons and calculated and plot the nullclines and fixed point(s) of the system. • The coding was done in Python (programming language). I have further explored by varying the network parameters to show different behaviors of the system like changing the number of fixed points of the system representing a binary task with short-term memory.• I have used Euler’s numerical integration to solve for and plot the trajectory of the activity of the excitatory and inhibitory population, thus representing the binary task with short-term memory.• I worked on exploring the stability of the network and shifting the network activity towards a stable fixed point.• At a particular input current to the excitatory neurons the system showed bifurcation.• The different network activities show that even a simple 3-neuron network can show different behaviors resembling short-term memory earlier observed also in 2-neuron model. Show less

• 

  NASA MINDS

  Sept 2023 - now

  NASA MINDS Undergraduate Student Design Competition

  • It is a multi-semester undergraduate level challenge that provides funds to student and faculty teams to design and build prototypes for technologies needed in support of NASA’s Artemis mission.• I am leading the team for developing theoretical framework and algorithms of the “Swarm Robotics” project.• Our objective is to develop mathematical framework and simulation(code) for how individual members(robots) in the swarm coordinate with each other and move collectively as a group.• I am applying Particle Swarm Optimization (PSO) algorithm and Artificial Neural Network (ANN) algorithm.• PSO alone is computationally expensive. So, I used the ANN called Stochastic Gradient Descent (SGD) for fast optimization.• To increase accuracy, we divide the cost function into mini-batches.• I applied successive layers of mini-batch SGD• Working with a team of enthusiasts helped me learn to work in collaboration. Show less

• 

  Bunker Hill CC

  Jan 2024 - now

  Math Tutor

  Working as a Math Tutor in my college community teaching developmental math and high school and college mathematics. Helping students of different ages, returning students and underrepresented students.Has been a huge growth experience, developing skills like collaboration, compassion and confidence.

• 

  Massachusetts Institute of Technology

  Jan 2024 - Jan 2024

  Quantitative Methods Workshop (QMW), 2024 Student

  • One-week intensive workshop has been specifically designed to introduce students to quantitative tools and programming languages such as Python, which are routinely used to analyze experimental data in biology and neuroscience.• Lectures and hands-on coding tutorials.• Introductory Python programming, plotting and curve fitting in Python, Machine Learning applied to Cognitive Neuroscience, Brain Computer Interface, Signal Processing applied to neuroscience, Cryo EM, Cognitive Neuroscience, Memory, Genomics/GWAS/RNAseq/Bioinformatics.• Lunch and discussion with MIT faculty.• Prepares students for summer research and graduate school. Show less

• 

  Harvard Medical School

  Sept 2024 - now

  Research Intern

  Computational neuroscience research on Network mechanisms for brain computations like memory