Wearable EDA Quality Model improves SOTA
Published:
Improve Electrodermal activity (EDA) signal quality SOTA by $8\%$ by unsupervised pre-training
Blog posts
2025
Wearable EDA Quality Model improves SOTA
Published:
Electrodermal activity (EDA) is a key indicator of sympathetic nervous system activation and a reliable marker of emotional arousal or stress. However, motion artifacts and connectivity issues often degrade EDA signal quality. To enable meaningful interpretation, it is essential to distinguish between high- and low-quality EDA signals. We propose an EDA signal quality index system leveraging unsupervised pre-training—a strategy widely used in natural language processing models such as GPT. Our approach achieve approximately $8\%$ in ROCAUC improvement compared to SOTA, while requiring only half the training epochs. This demonstrates that even with limited labeled data and a lightweight model, pre-training can significantly enhance EDA quality assessment, making it practical for real-time, wearable health applications.
2024
Multi-task learning and Multi-Modal learning
Published:
Multi-task Learning solves multiple tasks at one time, while exploiting commonalities and differences across tasks. Multi-modal Learning integrates and processes multiple types of data in one framework.
2023
Chat-GPT assists programming and debugging
Published:
How Chat-gpt from Open-AI going to affect our day-to-day work?
2022
Tutorial for audio MNIST dataset
Published:
This is a tutorial for analyzing audio MNIST (digits) dataset
notes for deep one-class classification
Published:
Notes for reading paper: learning and evaluating representations for deep one-class classification link.
Predict unit-level LoS using time-series modelling
Published:
We propose to a dynamic, real-time, lightweight, end-to-end next day median ED-LoS prediction algorithm using time-series LSTM modelling. Similar method can be extended to unit-level KPI prediction.
Challenges in data labels
Published:
I here list some different challenges in data labels that I encountered in the past and ways to model them.
Reverse gradient searching for turning point
Published:
This note summerizes using a simple and efficient trick: reverse gradient searching for finding turning points in noisy cases.
2021
Time series clustering using tslearn
Published:
TSlearn package for classic timeseries clustering methods.
Notes from reading papers on SBM and GNN
Published:
Notes from reading papers about Stochastic Block Models and Graph Neural Networks
2020
Automatic python script start setup
Published:
A simple tutorial on how to host a python service on a server using bat script and task scheduler.
OMP for sparse signal recovery
Published:
Ortogonal Matching Pursuit for recovering sparse signals