I am a MS student in Computational Biology and Biomedical Informatics at Yale, with a BS in Biomedical Engineering

Qinyao Hou  • Peerlist

- Built a real-time assistive navigation system on Raspberry Pi integrating LiDAR and ultrasonic sensors for environmental perception - Developed sensor fusion algorithms to detect obstacles and terrain changes within ~1.5 m using depth map processing and distance sensing - Implemented Python-based real-time detection pipeline for hazard identification and decision logic - Engineered dual-motor haptic feedback system to deliver directional guidance based on obstacle position - Optimized system latency and responsiveness for continuous real-world operation - Designed for assistive navigation applications for combined visually and motor impaired users

Assistive Navigation System using LiDAR, Ultr

- Designed and compared multiple neural network architectures (1-layer, 2-layer, and 3-layer models) - Evaluated model performance across different text representation techniques - Conducted systematic comparison using accuracy metrics to assess model effectiveness - Analyzed trade-offs between model complexity and predictive performance

Neural Network Architecture Comparison

- Processed 1000+ text reviews using TF-IDF vectorization with unigrams and bigrams - Applied preprocessing techniques including lemmatization and stopword removal - Trained and evaluated multiple classification models: Naive Bayes, Logistic Regression, SVM, and Random Forest - Compared model performance across different feature representations and reported accuracy metrics Built end-to-end pipeline for text preprocessing, feature extraction, and classification

Text Classification using TF-IDF and Machine 

- Implemented LDA topic modeling on large-scale text datasets (1000+ reviews) - Extracted latent topic distributions and identified top contributing terms per topic - Analyzed patterns across different domains (e.g., restaurant vs. movie reviews) - Interpreted probabilistic outputs to generate meaningful insights from unstructured text data