Engineer Intern Machine Learning

Steven Tran’s Resumé *

EDUCATION, EXTRA CURRICULUM ACTIVITY & LANGUAGES

Vietnam National University Ho Chi Minh City – University of Science

• Faculty: Bachelor of Mathematics & Computer Science

• Major: Data Science

• Relevant coursework: Programming Fundamentals, Data Structures and Algorithms, Computer Networks, Data Mining, Introduction to Machine Learning, … 2020 - Ongoing

Ho Chi Minh City – Bui Thi Xuan High School

Academic achievement:

• Three (03) years with good grade overall

• Second place in the City-level Chemistry competition for excellent students in grade 12 and grade 9

• Excellent score in Math and Chemistry in National Highschool Exam (9.4 and 9.5) 2017-2020

E/C Activity: Member of Science and Technology Club (CLB Khoa Học Kĩ Thuật) 2019-2020 Languages: Vietnamese (Native); English (Advanced proficiency – C1) EXPERIENCES WITH PROJECTS

FILM REVIEWS SUMMARY Personal Project Ongoing

• Proficient in RNN and Transformer-based models, well-versed in various tokenization techniques, and experienced in preprocessing and cleaning textual data for effective processing.

• Developed a GUI app utilizing DNN (Dense Neural Network), RNN (Recurrent Neural Network) and Transformer Neural Network. Employed word2vec for vectorizing the entire text in the DNN model and vectorized each word individually using word2vec in the RNN model.

• Expanded the app's functionality by incorporating BERT-tiny, a Transformer-based model, enabling prediction of review ratings on a scale of 1-5. This enhancement surpassed the binary positive/negative prediction capability of the previous models. Gained comprehensive understanding of BERT's inner workings, ensuring effective utilization within the application.

AGE and GENDER PREDICTION PROJECT Personal Project 04/2023

• Built a simple GUI webcam app

• Gained proficiency in deep learning, transfer learning, and TensorFlow, including the ability to modify layers in a pretrained model. Demonstrated a thorough understanding of FaceNet, ArcFace, VGGFace, and other models, as well as the use of transfer learning with these models

Project Outcome: Achieved a real-life gender accuracy of approximately 95% and age MAE under 2.5. To attain these performance metrics, the optimal image cropping area needs to be determined through testing, as the model is trained on a dataset with faces that are cropped similarly. Built an app to record the optimal cropping area for an image and incorporate it into the webcam app's GUI. However, it should be noted that these results have not been widely validated and have only been verified through testing on images collected from the internet and family/friends BIGINT CALCULATOR Personal Project 01/2023

• Created a program that can calculate big numbers up to 1000 digits very fast (basic math), and can determine whether the number could be a prime number using multiprocessing in C++ and Python with reasonable time (with 183 digits number: ~500s in Python and 40s in C++)

• Implemented object-oriented programming concepts in both C++ and Python AMERICAN SIGN LANGUAGE DETECTION PROJECT Team of 4 11/2022

• Led the team of 4, organized and delegated tasks for members, monitored the overall progress

• Built model and GUI app using TKinter

MINH TRAN LE, STEVEN

District 4, HCMC /StevenTran0410 039.303.4871 ****************@*****.*** I'm hoping to land an Machine Learning/AI Engineer intern position where I can put my coding and data analysis skills to use while gaining real-world experience in a work setting. My love of coding extends to problem-solving in the area of computer science, and I am constantly keen to learn and keep up with new developments in science and technology. Steven Tran’s Resumé 2

• Developed computer vision models using OpenCV and MediaPipe to recognize American Sign Language gestures

Project Outcome: Real-life accuracy was around 70% for people who were just taught how to use sign language, and reached up to 90% for those who used sign language well already (Even though it has not been thoroughly validated yet, through some tests with university fellows, the figures were as mentioned above.)

INSURANCE PRICE PREDICTION Team of 3 04/2022

• Contributed to data cleaning and pre-processing, and visualized data using Python libraries such as Pandas and Matplotlib

• Developed machine learning models to predict insurance prices using Scikit-learn TECHNICAL SKILLS

• Programming languages:

Advanced: Python, C/C++

Intermediate: Matlab, R, SQL

• Tools: OpenCV, TensorFlow, Keras, PyTorch, and other libs/models that are common for Machine Learning or Deep Learning

• Familiar with Object-Oriented Programming, Multiprocessing programming (in Python and C++)

• Familiar with cloud services GCP (Through #QuanquanGCC)

• Familiar with Docker

SOFT SKILLS

• Technology-orientation

• Problem-solving

• Logical thinking

• Adaptability

• Result-orientation

Contact this candidate