Selected Projects

Google Drive for the slide:
https://drive.google.com/drive/folders/1a7zb4ilW0Bfinz4Gm2YI_pzsRCRgJlYD?usp=drive_link

1. 2026 I have come back to China from US, attending an academic project at Tsinghua University focusing using bio-inspired algorithm like hebbian learning to enhance the classical machine learning framework, and using hyperbolic topology to discover the latent geometric shape under the biological brain.

2. I was working as a fellow at Harvard, using reinforcement learning and recurrent networks to simulate zebrafish collective behavior. We are planning to map the artificial neural network onto the biological neural network focusing on the whole brain dynamic. The code is here: https://github.com/ding9858/RL_for_Zebrafish

3. I completed my academic visitor work "Simulation based inference for MRI diffusion models" at Oxford. I implemented different simulation-based inference methods for MRI diffusion models, then used the model output to track the connectome of the human brain. The final report is named "Report:Simulation_based_inference_for_MRI_Diffusion_models.pdf" in the google drive link. The code is here: https://github.com/ding9858/SBI_MRI_diffusion

4. I have attended some seminars at the Max Planck Institute of Biological Intelligence and wrote an overview of electron microscopy connectomes, including the entire pipeline from data acquisition and storage to segmentation details. The final report, titled 'Report: Electron Microscopy for Connectomes Segmentation', can be found in the google drive link.

5. I did an internship at TUM that analyzed PET brain connectivity using the FSL tool package. This includes white matter fiber tractography based on PET images. Registration is performed on T1/PET/resting state/etc., after which the standard brain atlas is mapped onto our data for analysis of brain region connectivity. The final report, named 'Report:FSL_for_Connectomics.pdf', can be found in the google drive link. The GitHub code link is: https://github.com/ding9858/FSL-brain-conenctivity-

6. I attended a practical project at TUM, during which I developed a deep learning model based on EEG data to classify neural pain. The model uses a self-supervised learning (contrastive learning) pipeline to improve accuracy in classifying disorders. The final report is named 'Report: EEG_Selfsupervised_learning_Pain_Classification.pdf' in the google drive link. The code link is: https://github.com/ding9858/Self_Supervised_Learning_on_EEG

7. I have attended a practical course at TUM, in which we have trained a model using self-supervised multimodal learning based on different MRI data like T1,T2,MRA etc. which is potentially useful for Alzheimer's classification and connectomes analysis. The final report is named "Report:MRI_Multi_Modalities_for_Neuroimaging.pdf" in the google drive link. The GitHub link for the code is: https://github.com/sven-luepke/mlmi-project

8. I have cooperated with Oxford, Harvard and TUM, wrote an research proposal at Max Planck Institute about using the intepretable machine learning and statistics inference to research the multi modal brain connectome, integratin tthe molecu and cellular information. 

Undergraduate Projects

1.2019 March - 2019 October, Germany, Göttingen, University of Göttingen

Tutor (Group of Dr. Henrik Brosenne, Course: B.Inf.1801: Basics of programming with C ) Assist teaching the basics of programming with C, like standard data types, constants and variables, pointers, strings, fields and stacks. 

2.2019 May - 2019 September, Germany, Göttingen, University of Göttingen

Student research assistant(Dr. Tomas Kulvicius, “Path planning for maze navigation using planning operators”. )The goal of this internship was to improve a path planning algorithm for maze navigation

3.2020 May – 2020 September, Germany, Göttingen, University of Göttingen 

Bachelor thesis (Group of Prof. Dr. Xiaoming Fu. “Imbalance data set problem in the training of neural network”.) The goal of this research is to solve the problem of imbalance date set in different convolutional neural networks. 

Please find the code, the slide and my profile in the link of github, linkedin, and google drive bellow.