*

Amin Jabini

Los Angeles, CA ***** Phone: 213-***-**** Email: adkdat@r.postjobfree.com Linkedin

Education

University of Southern California, Los Angeles, USA

• PhD, Engineering Jan 2019-Dec 2023

Annenberg Fellowship Award (2019)

• MSc, Computer Science Jan 2020-Dec 2021

Sharif University of Technology, Tehran, Iran

• MSc, Engineering Aug 2015-Jan 2018

• BSc, Engineering Aug 2011-May 2015

Graduated Summa Cum Laude

Relevant Coursework

• Machine Learning (ongoing at USC) Analysis of Algorithms (at USC) Foundations of Artificial Intelligence (at USC) Deep Learning Specialization (on Coursera) Deep Reinforcement Learning (Online-UC Berkeley)

• Probability for Electrical and Computer Engineers (at USC) Uncertainty Quantification (at USC) Structural Reliability and Probabilistic Modeling (at SUT)

• Linear Algebra for Engineering (at USC) Advanced Engineering Mathematics (at SUT) Technical and Language Skills

• Programming: Python, MATLAB, R, C++

• Frameworks/Packages: Git, TensorFlow, Keras, SQL

• Proficient in probability and statistics

• Proficient in Word, Excel, PowerPoint, LaTeX

Projects

Research

• Graduate Research Assistant at USC Jan 2019-present o Proposed and coded a framework for sensor placement optimization under uncertainty. Based on Deep Reinforcement Learning o Implemented various Deep Reinforcement Learning agents using imitation learning, policy gradient and Double Q-Network for benchmark Open AI Gym control environments

o Formulated an efficient approach for quantification of sensitivity-based information gain in locally nonlinear systems.

• Research Scientist at INSURER Jan 2018-Dec 2018

(Center for Infrastructure Sustainability and Resilience Research) Implemented a Kalman-based input-state-parameter estimation framework for model identification in MATLAB. Achieved less than 10% final error in parameter identification using noisy sparse measurements with 80% initial error for a soil-structure system.

• Graduate Research Assistant at SUT 2016-2018

2

Analyzed known- and unknown-input Kalman filtering approaches to estimation of state and parameters in soil-structure systems for structural health monitoring purposes.

Other Projects

• Policy gradient algorithm for control problems

Implemented policy gradient algorithm with variance reduction technics, such as base-line correction and reward-to-go and applied to inverted pendulum, cartpole, and half-cheetah environments

• Actor-critic agent for Atari Games

Implemented Q-learning agent using for Pong Game and Lunar Lander.

• Sign Digit Classification using ResNet with Keras Using the SIGN dataset, implemented a 50-layer Resnet to classify the hand-sign digits and achieved 90% accuracy

• Car detection with YOLO algorithm

Coded a car detection model using YOLO algorithm to detect cars as well as the anchor boxes.

• Neural Style Transfer (NST)

Implemented NST using VGG-19 model as the encoder to produce artistic effects for images.

• AI agent for the game of Go

Developed a minimax agent with alpha-beta pruning and a Q-learning agent for 5 5 Go in Python and defeated random player (100%), greedy player (95%) and aggressive player (95%)

Leadership

• Department senator at Viterbi Graduate Student Association (VGSA) Sep 2020-Jan 2021 Chosen by department-wide election; Coordinated social, academic and career events for the department.

• Research mentor in Viterbi Summer Project Summer 2020 Supervised two first-year undergraduate students in a two-month summer program. Designed and led them through a project to “Compute the Information Gain in Dynamical Systems” by holding weekly meetings. The output is going to be presented as a poster at Viterbi poster session.

• Research mentor at Viterbi Summer Institute (VSI). Summer 2020 Mentored four incoming USC undergraduate students as a part of VSI. Led the team through the project, “Probabilistic Approach to Structural Life-cycle Analysis”. Thought programming with MATLAB and checked issues through two meetings per week. Results were presented in an online poster session to 100+ audience.

• Research mentor at Viterbi Summer Institute (VSI) Summer 2019 Led two incoming USC undergraduate students as a part of VSI. Led the project, “Impact of input amplitude on Performance of Kalman Filter”. Taught programming with MATLAB, held paper discussions and checked issues through meetings every day.

Contact this candidate