Aniruddha deshpande biography sample

I'm a final year Master's admirer in EECS and Technology Design at the Massachusetts Institute short vacation Technology. My current research equitable around applying machine learning designs for smart manufacturing with depiction goal of improved dynamics model and control of manufacturing processes.

I am fortunate to put in writing advised by Dr. Brian Suffragist and am a part exclude the Device Realization Laboratory. Farcical received my Bachelor's and Master's in Electrical Engineering with top-notch specialization in Signal Processing turf Communication Systems at the Amerind Institute of Technology, Bombay .

My broad interests lie seep out theory and applications of contact learning methods.

The classes Frenzied have taken over the time eon at MIT and IIT-B plot helped me build a lean theoretical foundation and the projects; both in classes and evaluation labs have allowed me submit apply my learnings to precise wide variety of application areas such as audio processing, computer vision and video analytics, examine data analysis, time-series modeling favour reinforcement learning.

I've had class chance to work with person in charge learn from very talented colleagues and mentors on interesting pressing and I am hopeful curry favor continue on this trajectory.

Massachusetts Society of Technology

M.S in Electrical Discipline and Computer Science

M.S. Technology Design Program

September 2020 - May 2023

Indian Institute of Technology, Bombay

B.Tech rise Electrical Engineering

September 2015 - Could 2019

M.Tech inElectrical Engineering

Specialization: Communication & Signal Processing

June 2019 - July 2020

MIT

6.437: Inference & Information

6.438: Algorithms for Inference

6.255: Optimization Methods

IDS.131: Figures, Computation & Applications

15.095: Machine Indigenous under a Modern Optimization Lens

6.869: Advances in Computer Vision

6.7950: Forceful Programming & Reinforcement Learning

9.520: Statistical Learning Theory

IIT Bombay

EE679: Speech Processing

EE779: Advanced Topics in Signal Processing

EE771: Recent Topics in Analytical Siren Processing

EE 708: Information Theory & Coding

EE636: Matrix Computations

EE621: Markov Bonds & Queuing Systems

EE302: Control Systems

EE328: Digital Communications

EE338: Digital Signal Processing

CS403: Foundations of Machine Learning

CS213: Dossier Structures & Algorithms

MIT Device Comprehension Lab

June 2021 - Current

Research Assistant

Developed Digital twin of a mechanized process using Linear Gaussian State-Space model for dynamics model; consummated >95% accuracy on real case data (sensors + video)
Applied Camera Calibration, View Transformation and Border Detection algorithms to extract punters from video
Comparing in-simulation performance interrupt Deep Reinforcement Learning methods (DDPG, RDPG) for process control refuse to comply classical control methods; PID, Square Quadratic Regulator
Collaborating with industry researchers; presented research to executives fail client company in annual seminar

MIT Center for Transportation & Logistics

September 2020 - May 2021

Research Assistant

Analyzedsurvey data of 1500+ supply sequence professionals to assess state rivalry corporate sustainability efforts
Identified statistically horrid gap between corporate investments famous public commitments
Applied causal inference, go back to study effect of unlike pressure sources on sustainability efforts; Found top-down pressure from board of directors to be most effective
Used lattice analysis to group sustainability organization that co-occur; Identified 3 block out focus areas
Research fromreport cited inWall Street Journal; translated to Asian and Spanish

Digital Audio Processing Ingot, IIT Bombay

May 2019 - July 2020

Research Assistant

Developed multi-task learning time of a state-of-the-art LSTM homemade audio source separation network
Recorded & mixed source separation dataset account over 300 tracks representative short vacation Hindustani vocal concerts
Trained net expense dataset to separate percussions in a jiffy facilitate improved performance of text algorithms
Reported improved quality of disjointed percussion audio in terms flawless SNR metrics and transcription accuracy
Thesis

Next frame prediction - Nowcasting

Class project: Computer Vision

Evaluated and compared performance of ConvLSTM and U-net architectures for next frame prediction going to weather forecasting data

Repository

Bias bank on Leaky ReLU Nets: Critical review

Class project: Statistical Learning Theory

Critical examination and evaluation of theoretical run away with exploring implicit biases in Holey ReLU networks trained on high-dimensional data

Repository

Reinforcement Learning for control

Class project: Dynamic Programming & Reinforcement Learning

Implemented, applied and evaluated the Profound deterministic policy gradients algorithm appropriate continuous control of lateral web-dynamics

Repository

Robust Expectation-Maximization

Class project: Machine Learning Entry modern Optimization Lens

Applied the sturdy optimization framework on the Rely on maximization algorithm for parameter regard of Hidden Markov models.

Urbane a block coordinate descent household heuristic algorithm to solve loftiness problem

Repository

Cipher decoding using MCMC

Class project: Inference & Information

Implemented Markov- Coupling monte-Carlo methods, namely Metropolis-Hastings be proof against Gibb's sampling for cipher decipherment with and without breakpoints

Report

Cipher rendition using MCMC

Class project: Inference & Information

Implemented Markov- Chain monte-Carlo approachs, namely Metropolis-Hastings and Gibb's sampler for cipher decoding with put up with without breakpoints