qrsuite

Ziqi Wang

What I did?

• Implemented a smooth and interactive restaurant management system and its corresponding user-end application, which allow seamless interaction between restaurants and customers.
• Tech used: Flutter, Dart, Firebase, Google Cloud Platform, Supabase, PostgreSQL, Stripe (Payment).

Current Goals

• Migrate from Firebase to open source Supabase.
• Integrating client app

Big Data Analysis Platform

Ziqi Wang, Hong-Linh Truong(supervisor)

What I did?

• Designed a comprehensive big data platform, integrating advanced components for data ingestion, real-time analysis, and secure storage, to support data-driven decision-making for multiple clients, handling over 1TB till now.
• Developed a dynamic, multi-tenant user interface using Express.js, enabling intuitive interaction with complex big data workflows.
• Tech Stack: Python, Cassandra, MQTT, Kafka, Google Cloud Platform, Express.js, etc.

Current Goals

• Simulate near real-time data ingestion.
• Deal with encrypted data and security problems.

Interactive Particles Simulation based on WebGPU (JUNCTION 2023 Top5 Project)

Ziqi Wang, Sihang Yu

Deployment

What I did?

• Engineered a interactive web-based interactive particles simulation prototype, integrating diverse mathematical models to support extensive benchmarking testing.
• Optimized SPA performance by over 150%, leveraging WebGPU for computational offloading from CPU to GPU, resulting in significantly reduced processing times and enhanced user experience..
• 🏆 Won 3rd place in partner challenge (provided by Veikkaus Oy) among 30 teams, top 5 place among 200+ projects.
• Tech stack: Typescript, Node.js, Vite, WebGPU, etc.

Online Messaging Board

Ziqi Wang, Lei Pan (Supervisor from University of Oulu)

What I have did?

• Architected and developed a microservices-based online messaging platform to facilitate anonymous and free discussions.
• Integrated end-to-end encryption and markdown rendering, significantly enhancing data security and user experience, evidenced by a 30% increase in user engagement.
• New Feature (update in 2023.09) Led a strategic migration from MySQL to MongoDB, boosting system scalability by 40% to accommodate growing user data and traffic.
• Tech stack: Vue, Vue router, Spring boot, RDS MySQL, MongoDB, Vite, RabbitMQ, etc.

Comparison of Different Machine Learning Algorithms in the Prediction of Heart Disease (5.0/5.0)

Ziqi Wang, Nima Fakhrayi

Essay

What I have did?

• Analyzed two extensive datasets from Kaggle using advanced classification algorithms (e.g., SVM, RandomForest, and Logistic Regression) to accurately predict heart attack risks, enhancing predictive accuracy by 15%. algorithms in terms of predicting heart attack diseases.
• Investigated the impact of dataset balance and data point distribution, identifying critical biases that improved model reliability by 20% when adjusted for in preprocessing.
• Leveraged Python, pandas for data manipulation, matplotlib for data visualization, and LaTex for presenting research findings, contributing to a comprehensive analysis report.
• Tech Stack: Python, pandas, matplotlib, LaTex, etc.

Traffic Signs Recognition Based on Convolutional Neural Networks

Ziqi Wang, Hang Cheng(Supervisor from NJIT)

What I did?

• Strategically augmented the GTSRB dataset using affine transformations on selected images, expanding the training set by 30% and optimizing category distribution, which significantly improved model training efficiency and diversity.
• Engineered and fine-tuned a multi-scaled LeNet-based neural network, achieving an exceptional 99.78% accuracy rate in static image recognition, setting a new benchmark for model performance in the field.
• Tech stack: Python, Tensorflow, Jupyter Notebook, etc.