About me
I am Hilda Ibriga, currently a PhD student in Statistics at Purdue University, working under the supervision of Prof. Bruce Craig and Prof. Wei Sun.
My research interests include the theoretical analysis of tensors and their application to machine learning. I am especially interested in developing tensor analysis methods that can be used to boost the performance of recommender systems, missing data recovery and personalized recommendations algorithms. To this end my PhD research is centered around the theory of tensors but also in developing tools to facilitate the integrations of tensor methods into pre-existing machine learning algorithms.
I have also worked on applied projects in reinforcement learning and would like to pursue more work in this field in the coming years.
I have three years of work experience as a statistics consultant during which period I had the opportunity to work on over 50 projects from beginning to completion. I find great joy in sharing the knowledge I have acquired over the years and I have been fortunate to do so through four years of teaching a wide variety of undergraduate and graduate level Statistics and Maths courses.
- Tensor Analysis
- Machine Learning
- Reinforcement Learning
- Statistical Inference
PhD in Statistics, 2021
Purdue University
MS in Mathematics, 2014
University of Arkansas
BSc in Mathematics & Economics (double major), 2011
Westminster College
Skills
Computing
Expertise
Language
Work Experience
Purdue Data Mine is the first large-scale living learning community for undergraduates from all majors, focused on Data Science for All.
Responsibilities include:
- Supervised a group of 12 TAs, organized and ran TA training meetings
- Contributed to writing, reviewing all R, python and SQL course projects
- Held office hours and led projects grading meetings
- Contributor and editor of the DataMine example book
Responsibilities include:
- Conducted comparative research on two competing model agnostic machine learning interpretability methods Lime and Anchor.
- Worked on integrating Lime into the existing machine learning model for predicting account churn which allowed to:
- Identify features which explain high churning probability for a given account
- facilitate understanding and actionability for the business team
- Co-wrote documentation for the implementation and integration of Lime into the account churning model
Responsibilities include:
- Worked on 50+ consulting projects on fields such as: Engineering, Social Sciences, Natural Sciences.
- Assisted clients at each of the fundamental statistical modeling steps:
- Defining scope of project, design of experiment,
- Data quality control, data analysis and visualization using R, SAS, MATLAB or SQL
- Results interpretation and writing for journal publication, technical report and grant proposal.
- Projects include social network analysis, sample size calculation for complex experimental designs, analysis of large time series data, metrics engineering.
- Co-authored the free manual titled “Introduction to the Statistical Software R”, to provide a quick introduction to R for the use of faculty and students at Purdue University.