Python Aws

SHAWN GONSALVES

Email: *************@*****.*** https://www.linkedin.com/in/shawngonsalves7 https://github.com/shawngonsalves Contact No: 760-***-****

EDUCATION

• Master Of Science, Computer Science, California State University, San Marcos. August 2019 - May 2021, GPA: 3.5

• Bachelor of Engineering, Computer Engineering, Mumbai University. July 2019 - May 2019

TECHNICAL SKILLS

ML Libraries: Numpy, Pandas, Tensorflow, Keras, TensorflowJS Programming Languages: Python, C, Java, Prolog

Tools: Anaconda Navigator, Eclipse, Unity, SWI-Prolog, Android Studio, Visual Studio, WireShark, Jupyter Notebooks, Git, GitHub. Cloud-Based Technologies: Amazon Web Services, Google Cloud Platform, Firebase PROJECTS

Data Modelling on Alzheimer Data using Genetic Algorithm, Differential Evolution and Binary Particle Swarm Optimization( Sept 2020 - Nov 2020)

Technology and libraries - Python, NumPy, Pandas

ML techniques: Multiple Linear Regression, Support Vector Machine, Artificial Neural Network.

• Evaluated the model on three ML algorithms using the concepts of Genetic Algorithm that creates generations of the model to track the fitness value of each row of the matrix in order to find the optimal model with highest prediction accuracy.

• Added a routine for a function that deletes all the invalid/junk values from the data and returns a sorted and rescaled version of the data.

• Implemented key features like Data Cleaning, Data preprocessing and Data Modelling on extracted Alzheimer Dataset. Face Detection and Image Analysis Using Amazon Rekognition(Oct 2020 - Dec 2020) Language and Technology - Python, Amazon Web Services AWS Services: AWS Rekognition, AWS Lambda, AWS StepFunction, Amazon S3, AWS Elastic Search, AWS CloudFormation.

• Researched on implementing AWS Rekognition APIs in conjunction with other AWS services to build a system that can apply pre-existing AWS Rekognition filters to filter out photos that meet specific criteria.

• Gained a good learning experience to practically apply Cloud services, understand how powerful the services are, and how to use them together in tandem.

Using Neural Nets and TensorFlow to detect the presence of Pneumonia in a Patient(May 2020 - Aug 2020) Technology- Python, Keras, Tensorflow, Tensorflow-JS

• Designed a Convolutional Neural Network model from scratch that detects the presence of pneumonia in a patient based on their frontal chest X-Ray images.

• Used Keras as a backend for Tensorflow for Data pre-processing, augmentation, and model creation.

• Experience in moving trained ML model into production by deploying it on a web server using Tensorflow-JS.

• Received scholarship in the same from California State University, San Marcos for successful implementation of the Research. Voice-Based Email for the Blind(June 2020)

Technology- Python

• Implemented Python automation using different libraries and modules to help the blind access the email using just their voice.

• Completely eliminated the use of keyboard and mouse as the system prompts the user with voice commands.

• Developed a feature in the system that reads out the latest unread email sent to the user. Detecting the presence of CoronaVirus in a patient using concepts of Deep Learning(Feb 2020 - May 2020) Technology- Python,Tensorflow, Tensorflow-JS

• Trained a deep learning model that takes X-Ray images of patients as input and predicts if the person suffers from CoronaVirus.

• This significantly reduces the time taken in analyzing the images by radiologists and thereby increasing the efficiency of detecting the presence of virus.

PUBLICATIONS

Artificial Intelligence and its related Application in the Processing of Natural Language http://ijircce.com/admin/main/storage/app/pdf/bmdolVbqZI3aCFB73aHCGWXbOMHtOOSdHWMZi3eM.pdf International Journal of Innovative Research in Computer and Communication Engineering(IJIRCCE) July 2018, Volume 6 Issue 7.

CERTIFICATIONS

• Coursera - Crash Course on Python

• Udacity - Introduction to Tensorflow for Deep Learning

• Coursera -Browser based Models with TensorFlow.js