This is a collection of data science projects that I have done. Feel free to reach out to me on my LinkedIn!

In this project, using a dataset of pet profiles consisting of image, text and tabular data, I set out to explore ways to help animal shelters improve adoption outcomes of their pets in order to tackle animal shelter overcrowding and animal euthanasia. Firstly I created a classification model to identify pets that would not get adopted within 3 months. Features were engineered from the tabular data given and from the text and image using deep learning techniques to get the image (CNN model) and text (Transformers BERT) in the form of a vector to feed into the model. Knowing the pets that are likely not to get adopted within 3 months allows the animal shelter to better allocate their limited resources. In addition, I have also created an image search tool to allow potential adopters to search for similar looking pets in the listing by uploading an image of a pet.

In this project, I have scraped data from two subreddits, PS5 and Xbox Series X. With the goal of creating an automated tagging system for a forum, I used the data from the reddit post to develop a classification model to distinguish between a PS5 and and Xbox post. Various NLP technques such as TF-IDF was used to clean data and extract features of the text to feed into the model. In addition, deep learning techniques was used to perform sentiment and emotion analysis on the various topics within each subreddit.

In this project, I have built a regression model to estimate the sale prices of Houses in Ames. Using a dataset with around 80 features, data analysis and visualisation was performed to better understand the relationship between the sale prices and each variables. Apart from generating a prediction using various regression models, the variables were also analysed to infer the strongest features that influence house prices.

This projects explores the trend of SAT/ACT scores over three years, from 2017 to 2019. I looked at the difference in score patterns across the different States in the USA and between the two test. I also examined the relationship between the test and racial diversity of the state, the HDI (Human Development Index) of each states and the population density of the states.