ÂÜÀòÂÒÂ×

Kuan-chang Huang is an engineering program manager at Apple. Kuan-chang has also worked as a data research intern at Yahoo, a machine learning intern at Cathay Life Insurance, and an AI engineer intern at IBM.

Kuan-chang Huang’s experience in data research began at Yahoo, where they developed an E-Commerce recommendation system using BERT based on media news headlines and abstracts. Kuan-chang also pre-trained and fine-tuned GPT-2 model to automatically generate slogans for recommended items. In addition, they worked on reading and writing large-scaled data formats on HDFS using PySpark.

At Cathay Life Insurance, Kuan-chang Huang established a fraud detection system using Neo4j, Graph Neural Networks (GNN), and Decision Tree models. Kuan-chang was able to increase accuracy by 2% by applying graph-related features to the original LightGBM model.

Lastly, during their internship at IBM, Kuan-chang Huang achieved image classification model accuracy of 99% and compressed model size to 1.5% using knowledge distillation techniques on detecting specific ear diseases.