Zhenyu Ruan

B.Sc., Computer Science & Data Science University of Sydney

Foundational year focused on programming, data science, and informatics; transferred to UniMelb to continue B.Sc.

B.Sc., Computing and Software Systems University of Melbourne

Coursework focus: AI / ML, software systems, computer security.

UAV Algorithm Intern Zhejiang Public Information Industry · UAV Application Tech Research Institute

Contributed to core UAV control algorithms (flight stability, route planning, sensor fusion). Improved response speed and path accuracy of low-altitude inspection tasks by ~15% through PID parameter tuning and trajectory-planning refinement. Owned multi-scale aerial-image calibration, sensor data cleaning & fusion, plus end-to-end module integration for flight control + data acquisition + task-instruction pipelines.

AI Algorithm Intern Hangzhou Hikvision · Software Development Dept.

Built training frameworks for object detection (YOLO / SSD) and multi-object tracking (DeepSORT). Owned the data-preprocessing pipeline (annotation, cleaning, augmentation, label-consistency), developed the NMS & tracking-ID association module, and designed modular interfaces for the detection-tracking-classification pipeline. Shipped model export tools and an automated testing framework; familiarised with Hikvision's AI Open Platform architecture.

Data Platform Intern Big Data Development & Management Center · Shangyu District, Shaoxing

Contributed to the public data platform: data-resource integration, system-function optimisation & testing, and performance evaluation tracking for informatisation projects. Also worked on OA paperless-office software — requirements research, functional design, and module implementation & debugging.

Algorithms in Action — Interactive Sorting Visualization Platform

[Java · Gradle · Git · Figma · Confluence]

Capstone IT-project course with a 5-member agile team enhancing Algorithms in Action, an interactive platform for teaching algorithm design and performance. Designed and implemented the unified Stack Panel interface for sorting-algorithm visualisation, ensuring consistent animation and interaction standards across Quicksort, Merge Sort, and Radix Exchange. Coordinated UI behaviour, pseudo-code integration, and explanation text across modules with the Insertion / Selection team. Prepared three alternative Stack Panel versions, presented to the client, collected feedback, and finalised the unified display spec. Managed Git version control and maintained Confluence documentation for transparency & reproducibility.

Crime Rate Analysis across Victorian LGAs

[Python · pandas · scikit-learn · ML]

Built an integrated dataset combining socio-economic (IRSD, income, unemployment), demographic (education attainment, lone-parent families), and environmental (remoteness, housing, healthcare access) factors. EDA via correlation heatmaps, geographic crime plots, and offence-trend analysis. Trained and compared ML models (linear & ridge regression, random forest, K-means) — finding remoteness and socio-economic disadvantage as the strongest predictors. Delivered a 4000+ word analytical report with a reproducible codebase supporting policy recommendations.

Java Multiplayer Tank Battle Game

[Java · Processing · Gradle · JUnit · JaCoCo]

2D multiplayer tank-battle game with multi-level maps (28×20 terrain grid with dynamic destruction), turn-based controls, real-time scoring, projectile physics under gravity/wind, explosion area-damage, power-ups, and persistent scoreboards across levels. OO design with class-hierarchy + UML documentation; JUnit test suite reaching 90%+ coverage via JaCoCo.

Online Exam System (CLI)

[Python · CLI · modular design]

Python-based CLI examination system. Supports multiple question types (MC, fill-in-the-blank) with input validation and instant feedback. Instructor flow for creating & assigning exams; time limits, progress tracking, secure access controls; automated marking. Refactor cycle reduced redundant code by ~30% via modular programming.

Road Fatalities Data Analysis & Prediction (AU + NZ)

[Python · pandas · matplotlib · scikit-learn]

Three-stage Python project analysing road fatalities across Australia and New Zealand. Stage 1: collected and cleaned NSW / VIC / NZ datasets and built an integrated dataset with metadata. Stage 2: EDA on crash frequency vs. speed limits, time of day, road type, and weekday/weekend patterns; produced grouped aggregates and multi-attribute visualisations. Stage 3: trained linear-regression models to estimate cross-continent (Europe / Asia / Americas) death rates, evaluated via MSE / R². Delivered an analytical report with evidence-based traffic-safety recommendations.