Harish Ganesan's repositories
Music-Mood-Classification
This project classifies music into various moods, using several musical features.
News-Article-Classification-using-Spark
Built a data pipeline in Spark using MLlib to classify news articles into different categories with accuracy 94% among 3 models :
Reliable-Transport-Protocols
Implemented Alternating Bit Protocol and Go Back N Protocol
Autoencoder-MNIST
Classifying hand written digits from MNIST dataset using autoencoders in a deep network.
Automatic-Robot-Design-MCDPL
Building a land based rover with a catalogue of 4 main types of parts : Batteries, Chassis, Robotic Arm, Motor.
EDA-using-CDC-data
In this project, we create 5 charts in R using ggplot2 package, using data from the CDC website regarding the Flu Epidemic in the United States.
Laplacian-Blob-Detector
A program to generate blobs around areas of similar colour, texture, etc in an image
Street-Sign-Recognition-System
Built a model that classifies 5 different types of traffic-signs with 91.2% accuracy using MATLAB.
SummaRuNNer
The PyTorch Implementation of SummaRuNNer
WordCloud-from-Hadoop
Created word clouds in d3.js by using articles from New York Times and tweets from 3 topics using Hadoop Map-Reduce to analyze word frequency usage and word co-occurance
Celeb-Face-Classification-Tensorflow
Classification on Celebrity Face dataset into two classes - those who wear glasses and those who do not.
Flu-Analysis-in-R
Using twitter data, we map out the areas within the continental USA which have higher number of mentions of flu related topics. We then compare this to the actual flu map of the CDC and compare the two, and check if our hypothesis was validated.
Library-Management-OOP
A fun school project I did for managing books and accounts in a Library. Done using OOP concepts in C++.
Logistic-Regression-and-SVM-on-MNIST
This project aims to classify handwritten digits from the MNIST dataset using Logistic Regression as well as Support Vector Machines.
MediumEditor
MediumEditor using Materialize
MNIST-Classification-NeuralNets
This project classifies handwritten digits from the MNIST dataset, using Neural Networks.
Peer-Grading-System
This is a peer grading system with UI written in Java Swing, and unit test cases.
Regression-in-Python
Implementation of various types of Regression natively on a sample dataset.
Robot-Localization
Project using Robot data to localize the robot's current position.
Scene-Classification-Model
A model to classify 8 different scenes, namely: library, computer room, art gallery, garden, ice skating, mountain, ocean and tennis court
Stitching-Images
Stitch two images together using RANSAC for Homography and fundamental matrix estimation.