Shi Gengtian

Shi Gengtian

Data Scientist

Introduction

I’m deeply passionate about computer vision, natural language processing, and robotics. By day, I delve into algorithms and models to glean insights from visual data and bridge the gap between human and machine understanding. By night, I find solace in strumming my guitar, where creativity knows no bounds. Join me on this journey of exploration, innovation, and harmonious endeavors.

Education

Master of Science in Computer Science
Waseda University, Tokyo, Japan
2018 - 2020

Skills

  • Computer Vision: Object detection, image classification, and segmentation using OpenCV, TensorFlow, and PyTorch.
  • NLP: Sentiment analysis, text classification, and language generation with NLTK, spaCy, and Transformers.
  • Robotics: Motion planning, localization, and human-robot interaction using ROS, Python, and C++.
  • Machine Learning: Supervised/unsupervised learning, deep learning, and reinforcement learning.
  • Programming: Proficient in Python, experienced in C++ and Java.
  • Data Visualization: Matplotlib, Seaborn, Plotly for creating insightful visualizations.
  • Software Development: Git, agile methodologies, and CI/CD pipelines.
  • Problem Solving: Analytical and effective at dissecting complex problems.
  • Communication: Clear verbal and written communication skills.
  • Continuous Learning: Committed to staying updated with emerging technologies.

Talks

  • tbd

Kaggle

  • HuBMAP - Hacking the Human Vasculature: Ranked 84th out of 1,064 participants.
  • Feedback Prize - Learning Equality - Curriculum Recommendations: Ranked 69th out of 1,057 participants.
  • Feedback Prize - English Language Learning 3rd: Ranked 62nd out of 2,654 participants.
  • RSNA 2022 Cervical Spine Fracture Detection: Ranked 50th out of 883 participants.
  • HuBMAP - Hacking the Kidney (Image Processing): Ranked 51st out of 1,200 participants.
  • HM Personalized Fashion Recommendations (Recommender System): Ranked 53rd out of 2,952 participants.
  • Feedback Prize - Evaluating Student Writing (NLP): Ranked 63rd out of 2,058 participants.
  • RANZCR CLiP - Catheter and Line Position Challenge (Image Processing): Ranked 25th out of 1,547 participants.
  • Shopee - Price Match Guarantee (Multi-modal): Ranked 87th out of 2,426 participants.
  • HuBMAP + HPA - Hacking the Human Body: Ranked 53rd out of 1,175 participants.
  • BirdCLEF 2022 (Audio Signal Processing): Ranked 64th out of 807 participants.
  • CommonLit Readability Prize (NLP): Ranked 345th out of 3,633 participants.
  • TensorFlow - Help Protect the Great Barrier Reef (Image Processing): Ranked 132nd out of 2,025 participants.

Publication

  • Koshimizu, Takashi, et al. “Multi-dimensional affinity propagation clustering applying a machine learning in 5G-cellular V2X.” IEEE Access 8 (2020): 94560-94574.
  • Gengtian, Shi, et al. “Power control based on multi-agent deep q network for d2d communication.” 2020 International Conference on Artificial Intelligence in Information and Communication (ICAIIC). IEEE, 2020.