Portfolio of Ya Ting Hu
I’m a Data Scientist with a passion for turning complex data into practical, real-world solutions. My background is rooted in machine learning, natural language processing (NLP), and the growing field of generative AI (GenAI). Lately, I’ve been working a lot with large language models (LLMs), prompt engineering, and fine-tuning foundation models to help businesses innovate faster. I enjoy building models that not only predict outcomes but also create opportunities — whether that’s improving decision-making, streamlining operations, or developing entirely new products. Along the way, I stay focused on Responsible AI practices, ensuring that the solutions I deliver are ethical, explainable, and built for the real world. I’m always exploring new tools, techniques, and ideas, and I love collaborating with teams who see data not just as information, but as a strategic advantage.
Industry Solutions is a global organization comprising over 16,000 strategic sellers, industry experts, elite engineers, and world-class architects, consultants, and delivery experts. Together, we bring Microsoft’s mission of empowerment—and cutting-edge technology—to life for our customers and partners, driving value across their digital transformation journeys.
At Philips, we strive to improve people’s health and well-being through meaningful innovation, with a goal of enhancing 2.5 billion lives annually by 2030, including 400 million in underserved communities.
At ASW, we aspire to be the world’s leading health, beauty, and lifestyle retailer, striving to deliver more to our customers, colleagues, and communities.
Department of Mathematics and Computer Science
Department of the Built Environment
2020-2022 M.Sc. in ICT Innovation (GPA 9/10)Taken Courses
| ||
2020-2022 M.Sc. in Computer Science and Engineering (GPA 9/10)Taken Courses
Extracurricular Activities
| ||
B.Sc. Data Science (Cum Laude)Taken Courses
Extracurricular Activities
|
Five small projects, namely finding textually similar documents, discovery of frequent itemsets and assocation rules, mining data streams, graph spectra, and k-way graph partitioning using JaBeJa
Analysis of Den Bosch waste water treatment plant
American Airlines Twitter team performance analysis
An overview-based interactive visualization tool for temporally long dynamic graph sequences is implemented and applied on US domestic flight data.
The goal of this project was to predict the remaining cycle time of a request into an administrative system. Different models were designed and tested on the datasets provided by the BPI Challenges 2012 and 2019.
A bracelet creating the overall festival experience by connecting to stage lighting and guiding you to the bar with shortest line.
In this project we use the Spark platform for processing massive data and discovering minimal non-trivial functional dependencies and (soft) functional dependencies.
Two labs regarding HDFS, HBase, Hadoop, MapReduce, Spark, Spark SQL, Spark Streaming, Structured Streaming, GraphX and project regarding real-time Bitcoin price predictions with news sentiment analysis
HoodFood, a comparison tool which allows customers to compare prices between groceries easily, started by Diverse IT