projects
Data Arcade Tournament 2022 – 1st Place (MOE)
Achieved 1st place within MOE for the Data Arcade Tournament, which is an annual visual analytics competition exclusive for officers serving in the Singapore Public Service.
Maximising Profitability From Hotel Bookings
Provided a data science solution to a real-life business scenario of predicting hotel booking cancellations. Machine learning models of clustering and classification were used to classify groups of clientele with high risk of cancellation, as well as identifying top cancellation reasons. From these, a proposed business model was drafted to help hotels maximise their revenue from hotel bookings.
Secret Recipe Network
By studying a network of recipes, their attributes as well as user interactions and reviews with these recipes using PageRank, Degree Centrality and Similarity Ratio, top users and food communities can be identified in order for a food-based client to explore potential marketing partnerships.
Fraud Detection in Credit Card Transactions
By using Density Based Spatial Clustering of Applications with Noise (DBSCAN) on a credit card transaction dataset, transaction outliers can be detected and flagged out as potential fraud cases for further analysis.
Money no enough – ANALYSING CPI CHANGES
Analysed SingStat data on average retail prices to Consumer Price Index changes across 2012 to 2021. Descriptive, correlative and predictive analysis was conducted on the dataset, culminating in an RShiny app embedded on our website to visualise the changes effectively.
Analysis of Singapore Paliarmentary Debates
Singapore’s Parliarmentary Debate data was crawled from the internet and Natural Language Processing was performed on the dataset in order to identify key topics most discussed over the last 5 years. Topic Modelling and Information Extraction was performed on the dataset using LDA, BERTopic and spaCy.
Analysing DEMAND FOR ELDERCARE FACILITIES
As Singapore’s population grows older, it is important to identify high-risk planning areas in Singapore with insufficient eldercare facilities. By employing Regression Analysis and Predictive Analytics, we forecast the future demand for eldercare facilities in each planning area, neatly showcased in an interactive RShiny dashboard.