Siyuan Wang - E-portfolio

Introduction

I am Siyuan Wang, currently a combined Bachelor's and Master's degree student (major Biomedical Engineering) at Southeast University, and on an exchange program at KTH (Royal Institute of Technology).

You can find my previous projects and skills in this e-portfolio.

project

Project about AI

1.EmoDB dataset with SVM emotion classification -> machine learning

The main task of this project was to classify emotions in speech.

I built an SVM machine learning model to classify the IS09 features from the EmoDB dataset, achieving good classification results.

Result of the project, accuracy more than 75% :

Core code using Python, mainly about the model I used:

MNIST handwritten digit dataset image recognition -> deep learning

The main task of this project was to recognize handwritten images in the MNIST dataset.

Designing a deep learning model to classify the MNIST dataset using Pytorch and achieve very high accuracy.

Result, more than 98%:

Model code:

Core code:

Project about Signal processing using MATLAB

The primary focus was on basic filtering operations to remove artifacts and noise, facilitating signal analysis.

Operations were performed using MATLAB as the tool.

Only a subset of the operations that can be performed using MATLAB will be demonstrated.

1.Wavelet transformation for filtering.

Result, remove 50Hz noise:

2.Obtain the Hilbert spectrum.

Project about EEG signal processing using MATLAB

The project mainly aims to identify differences between normal individuals and patients using EEG signals to identify brain damage in insomnia patients and assist in their treatment.

I preprocess, segment, and perform statistical analysis on EEG signals in MATLAB, and generate analytical charts for other researchers to examine.

(Due to confidentiality reasons, only a portion of the results will be presented.)