The Collection
Data science and machine learning projects, 16 cards in all. The rare pulls come first.
Cassava Leaf Disease Detection using Neural Networks
Deep learning model using transfer learning on EfficientNetB3 to classify cassava leaf diseases, reaching 89.90% accuracy.
Detection of Animal Species in Tropical Landscape
Audio classification of rainforest species using mel spectrograms, ResNet-50, and AWS-scale data processing.
A Machine Learning Approach to Portfolio Management
A Random Forest stock-picking model that returned +240.70% in a simulated 2016-2020 portfolio, versus a 12.29% PSEi loss.
Creating Monet-style Art using CycleGAN
Using CycleGAN to translate landscape photos into Monet-style paintings without paired training samples.
Airbnb Recommender System for New York City
Sentiment analysis of user reviews combined with user-based collaborative filtering to recommend NYC Airbnb listings.
Machine Learning
Big Data, Data Mining, and Unsupervised Clustering
- Land of the Rising Sun: Uncovering Clusters from Airbnb Listings in Tokyo, JapanNo. 009
- May The Force Be With You: Analyzing the Difference in Rating System between IMDB and Rotten TomatoesNo. 010
- Clustering Philippine Stocks based on Historical Price Movement: A Dynamic Approach to Stock ScreeningNo. 011
- Eyes on the Road! Uncovering Patterns in California's Traffic Collisions using Advanced Data MiningNo. 012
Natural Language Processing
Exploratory Data Analysis and Applied Computational Statistics
- Thank Heaven for 7-Eleven: A Sales Analysis of a 7-Eleven Store in Muntinlupa CityNo. 014
- Analysis on Customs Import Data: Manufacturing, Food, and Clothing as the Strongest Business OpportunitiesNo. 015
- A Global Warning: Assessing the Inconvenient Truth by Analyzing Global Surface Summary of the DayNo. 016