This page was created to showcase all the projects associated with the courses completed during my masters. This page serves as an index to all the projects that are hosted on GitHub. Feel free to fork this project to customize :)
I have chosen "COMPC" - Information Technology as my stream. It provides flexibility to choose the courses that I'm interested in. All my courses can be divided into 3 main categories - Research, Machine Learning and Software Engineering.
| Course | Course Code | Projects |
|---|---|---|
| Research Project C | COMP9993 | Coming Soon |
| Research Project A | COMP9991 | Coming Soon |
| Course | Course Code | Projects |
|---|---|---|
| Computer Vision | COMP9517 | 1. Implementing Connected Component Algorithm In this project, I learned basic operations like adjusting brightness, the contrast of the image and other basic operations of image processing. Mainly this project taught how to implement Iso-data intensity thresholding, generating a binary image from a given image and performing connected component analysis on the image. Links: Project Source Code | Report. 2. Classification of Ara and Tobacco Plants I used HSV colourspace to extract the green part of the plant, I used SIFT and HOG algorithms to extract features and an SVM Classifier to classify the plants. Precision, Recall and ROC were used to evaluate the results. You could read about the techniques and code here. Links: Report. 3. Detection and Segmentation of Ara and Tobacco Plants - Group Project There were 3 tasks in this project. Task 1 - To detect the plants in the image, Task 2- creating a binary image of the plant, Task 3- Multi-instance segmentation. I worked on Task 3. Links: Project Source Code | Report. |
| Neural Networks and Deep Learning | COMP9444 | 1. Japanese character recognition Implemented various Neural Networks to solve this problem. Experimented with various hyperparameters influence on each network. The results can be found in the report. Links: Project Source Code | Report. 2. Predicting the product review of products sold on Amazon. With all the constraints in place, implemented text preprocessing methods. Used LSTM with the linear neural network model and a regularizer tnn.dropout to design the network. Links: Project Source Code |
| Big Data Management | COMP9313 | 1. Collision Counting LSH (C2LSH) Algorithm Implemented C2LSH algorithm in PySpark without using any inbuilt functions. The implementation was one of the fastest in the class of 300. Links: Project Source Code | Report. 2. Stacking Technique in Machine Learning using PySpark. Implemented solution using a bag of words, meta-feature generation, SVM and Naive Bayes methods. Links: Project Source Code | Report. |
| Machine Learning and Data Mining | COMP9417 | 1. Implementation of Stochastic gradient descent Implemented Stochastic gradient descent in Python and used it while predicting the housing prices. Links: Project Source Code 2. Text classification to predict the most relevant news articles for 10 Topic. Used Bag of words to construct the feature vector. Evaluated Precision, Recall and F1 score of each Machine Learning Model. Also, used SMOT technique to handle the imbalance in the data. Links: Project Source Code | Report. |
| Data Services Engineering | COMP9321 | 1. Revenue and Rating prediction of the Movie dataset. Implemented 2 different models to predict Revenue and Rating. Considering the length of the feature vector was large, tried feature reduction techniques. However, given the environment, it was taking a long time to train on UNSW CSE Machine. Hence had to let go of the method. Links: Project Source Code | Report. |
| Course | Course Code | Projects |
|---|---|---|
| Software Service Design and Engineering | COMP9322 | 1.Building a dental clinic system using a chatbot and dockerized services Implemented an ML-based ChatBot to handle the Dental Clinic reservation system. Used Wit.Ai to handle the natural language processing. Containerized each of the services (Chatbot, Time-slot and Dentist Service) Links: Project Source Code |
| Web Data Compression and Search | COMP9319 | 1. Implementation of Arithmetic Encoder and Decoder Implemented this solution in C++. Passed all the test cases. Links: Project Source Code. 2. Searching & Decoding BWT Encoded DNA Sequence Implemented BWT backward search and a decoder that efficiently decodes BWT encoded DNA file back to their original DNA sequence. The challenge in this assignment was to implement it so efficiently that it beats the execution of binary provided by the professor. I did implement a solution that passed all the test cases and beat the professor code by a good margin. Links: Project Source Code. |
| Data Services Engineering | COMP9321 | 1. Data Service for World Bank Economic Indicators Developed a Flask-Restplus data service that allows a client to read and store some publicly available economic indicator data for countries around the world, and allow the consumers to access the data through a REST API. Links: Project Source Code |
| Data Structures and Algorithms | COMP9024 | 1. Sydney Transport Planner Implemented a C Program for finding an optimal train connection that takes into account a given arrival time and user preferences. Links: Project Source Code |
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent