ddhruvgupta Goto Github PK

Welcome to my profile. You can browse through some of the projects that I have worked on in my repositories.

⚡ Here is a hightlight of some of the different projects that I have worked on:

• IoT: Leveraged Bluetooth to identify if registered users are in the office, using a network of Raspberry Pi computers to ping the Bluetooth NIC at a 10 min. interval, update PHP Web Application with an created API used by a Native Android Application (Java) and Amazon Alexa (Node js)

• After School Program Management System: Designed website layout wireframes, built front end using BootStrap 4, jQuery for drop downs, Ajax for autocomplete and DataTables.js for interaction with data. Created module to accept signatures for a waiver module using Signature.js. Tested against SQL injection.

• Network Protocol Improvement: Analyzed and simulated algorithm for job dispatch and scheduling in edge networks. Identified bottlenecks in performance and proposed updates to algorithm.

• Smart home application using RF vulnerability: Etekcity wireless switches use unencrypted traffic over RF to control home lighting systems. Using in-built Python libraries, RF hardware and a raspberry pi it was possible to demonstrate that such home automation solutions could be attacked by a malicious actor.

• Worm for Ubuntu: Created a non weaponized worm program in C for Ubuntu that would infect a program used to create an RC4 cipher stream.

• Analysis of Chicago Crimes Dataset: Data mining project to carry out exploratory data analysis on a dataset of crimes in Chicago. Applied Principle component analysis to identify factors useful in predicting the kind of crime and used K means clustering to identify similarity in types of crime. Applied logistic regression to classify types of crimes, built models to predict type of crime using Logistic Regression, Random Forests and Support Vector Machines by leveraging the SciKit Learn package of Python.

• IoT: Use sensor data from mobile phones to infer active application. Used Principle Component Analysis to identify which sensors are important for task inference. Created models with Naïve Bayes, Random Forrest and Neural Networks to carryout inference. Leveraged SkLearn and TensorFlow for machine learning models.