A collection of Data Science/Software Engineering Projects for you to build your own portfolio
Machine Learning, Cloud Deployment, Container Technologies and more...
Question Answering & Sentiment Analysis
Python Web Application Deployment
3.5/5
About: For the current usecase a short paragraph has been taken from here https://www.supersummary.com/game-of-thrones/summary/ that has been used to illustrate the QA application. You can ask questions based on the the text input provided and the out-of-box HuggingFace model will answer the question. The sentiment analysis model has been trained on the Amazon Reviews dataset and embedded inside the application.
About:
Fake news is a nagging annoyance these days which can lead to the spread of misleading and fabricated information. The task of assessing the veracity of a news article is a complicated one.
However, a first step that can be taken in identifying fake news is to detect the stance of two pieces of text, i.e estimating the similarity of two pieces of text. In this project, you will be using NLP techniques to assess the similarity of a given news article's headline against the body of the article. You will be exploring a very popular Deep Learning Framework called Keras to run the project.
We have designed a custom Neural Network architecture to solve this problem, however please feel free to explore on your own.
Autocomplete: Next Word Prediction
Statistical NLP based application
3/5
About: This is a Python(Flask) based next-word prediction application that has been developed to deliver a workshop at Hack the North 2022(University of Waterloo). It uses a simple N-gram language model to build the text autocomplete functionality.
Natural Language Processing
Teach a machine to read
2.5/5
About: These are a collection of python scripts and datasets that will help you understand the basics of Natural Language Processing from text cleaning to creating word embeddings. All the datasets and corresponding python scripts are present in the GitHub repository.
About: This application is meant to teach you about Docker Compose. This project has been developed to deliver a workshop at Hack the North 2021(University of Waterloo). It is a very simple application that acts as a counter to count geese at the University(there are too many and I wouldn't dare to disturb them!). It uses Python and Racket programming languages under one single application and uses Docker to containerize the application in a microservice based architecture.
About: This project is aimed at enabling the traffic police to install minimum number of CC cameras in a given area with many intersections and streets. This is the classical optimization problem called Minimum Vertex Cover problem. In this project, the problem has been solved using three different approaches, using a SAT solver and two approximate algorithms implemented in C++ .