Machine Learning Engineer

Shruti Gadewar Email : *************@*****.***

www.linkedin.com/in/gadewarshruti Mobile : +1-213-***-**** Education

University of Southern California Los Angeles, CA, USA Master of Science in Electrical Engineering Aug. 2017 { May. 2019

SIES Graduate School of Technology Mumbai University,India Bachelor of Technology in Electronics and Telecommunications Jul. 2013 { Jun. 2017 Technical Skills

Programming Languages: Python, C, SQL, Latex

Tools: Matlab, Numpy, Scipy, Pandas, Scikit-Learn, OpenCV, Keras, Tensor

ow, PRAAT

Operating Systems: MAC, Windows, Ubuntu

Work Experience

Imaging Genetics Center Marina Del Rey, CA

Graduate Research Assistant May. 2018 - Apr. 2019

Large Scale Study of MDD using MRI scans from ENIGMA Consortium: Classi cation using Lasso, Logistic Regression and SVM to evaluate Major Depressive Disorder diagnosis based on brain-imaging measures and basic demographic information.

Technologies Used: Python, Scikit-learn, pandas

Projects

Cash Miner Game: Jan. 2019 { Apr. 2019

The goal is to nd the optimal control sequence for the virtual gure to obtain maximum expected rewards present in a grid using Markov Decision Process, Value and Policy Iteration.

Airport Gate Assignment: Jan. 2019 { Apr. 2019

The aim is to assign each plane a time to start landing and takeo procedure such that all the planes have been assigned landing strips, gates and takeo strips without any con

icts using constraint satisfaction.

Laser Checkmate Game: Jan. 2019 { Apr. 2019

The end goal is to win against the AI agent(opponent) in an adversarial game such that the opponent’s laser beam covers lesser area which has been implemented using di erent Search Algorithms: Greedy Search, Mini-Max with Depth Limited Search.

Speech Animation: Aug. 2018 { Nov. 2018

An experiment to automatically generate mouth texture animations to synchronize with the input speech by applying deep learning approach(Unidirectional LSTM with time delay) to speech features(Mel lter banks) and lip-markers extracted from the sequence of images.

Technologies Used: Python, tensor

ow, keras, openCV(dlib)

PUBG Finish Placement Prediction: Aug. 2018 { Nov. 2018 Predicted the placement of players in a given PUBG game with up to 100 players using Machine Learning methods like Lasso, Decision Tree Regressor, Random Forest Regressor, Adaboost Regressor and Light Gradient Boosting Method with best Mean Absolute Error of 0.0205.

Technologies Used: Python, pandas, scikit-learn, Light GBM

Pattern Recognition System: Jan. 2018 { Apr. 2018 Developed a Pattern Recognition System that works on given real world bank marketing data set which included data preprocessing, feature-space dimensionality reduction, cross validation, training and classi cation using Random Forest, SVM and Logistic Regression.

Technologies Used: Python, scikit-learn, pandas

Processing and Analysis of Human Voice for Assessment of Parkinson: Aug. 2016 { Mar. 2017 Using the symptoms of rough, shaky and breathy voice as the earliest indicators of Parkinson, disease was classi ed using various Machine Learning Algorithms like KNN, Random Forest, Decision Trees, SVM and ADABOOST with best accuracy of 94%.

Technologies Used: PRAAT, python, scikit-learn, pandas Publications and Achievements

IUI Conference 2019: "Towards Human-Guided Machine Learning"

ISTE Journal: "Processing and Analysis of Human Voice for Assessment of Parkinson Disease" (ISBN. 978-93-86171-02-3).

Contact this candidate