Yihao Li

Mr. Yihao Li

orcid

Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
26 March 2021 in Atmospheric Measurement Techniques
Verified
Journal Article
Open access
PDF file
RCRunlong Cai
Runlong Cai
YLYihao Li
Yihao Li
YCYohann CLEMENT
Yohann CLEMENT
DLDandan Li
Dandan Li
CDClément Dubois
Clément Dubois
MRMatthieu Riva
Matthieu Riva
The Orbitrap mass spectrometer has recently been proved to be a powerful instrument to accurately measure gas-phase and particle-phase organic compounds with a greater mass resolving power than other widely used online mass spectrometers in atmospheric sciences. We develop an open-source software tool (Orbitool, https://orbitrap.catalyse.cnrs.fr, last access: 4 February 2021) to facilitate the analysis of long-term online Orbitrap data. Orbitool can average long-term data while improving the mass accuracy by re-calibrating each mass spectrum, assign molecular formulae of compounds and their isotopes to measured signals, and export time series and mass defect plots. The noise reduction procedure in Orbitool can separate signal peaks from noise and reduce the computational and storage expenses. Chemical ionization Orbitrap data from laboratory experiments on ozonolysis of monoterpenes and ambient measurements in urban Shanghai were used to test Orbitool. For the test dataset, the average mass accuracy was improved from <2 to <0.5 ppm by mass calibrating each spectrum. The denoising procedure removed 97 % of the noise peaks from a spectrum averaged for 30 min while maintaining the signal peaks, substantially helping the automatic assignment of unknown species. To illustrate the capabilities of Orbitool, we used the most challenging and complex dataset we have collected so far, which consists of ambient gas-phase measurements in urban Shanghai. These tests showed that Orbitool was able to automatically assign hundreds of molecular formulae as well as their isotopes with high accuracy.

5 May 2023 in Applied Sciences
Verified
Journal Article
Open access
PDF file
YLYihao Li
Yihao Li
PYPing Yi
Ping Yi
For bulk commodity, stock, and e-commerce platforms, it is necessary to detect anomalous behavior for the security of users and platforms. Anomaly-detection methods currently used on these platforms train a model for each user since different users have different habits. However, the model cannot be trained adequately due to insufficient individual user behavior data. In this study, to utilize information between users and avoid underfitting, we propose a contrastive learning framework to train a complete global model (GM) for anomaly detection in a trading platform. By confusing the data between different users to generate negative samples, the model can learn the differences between users by contrastive learning. To reduce the need for individual user behavior data, this framework uses a GM instead of a model for each user to learn similarities between users. Experiments on four datasets show that models trained using our framework achieve better area-under-the-curve (AUC) scores than do the original models, proving that contrastive learning and GM are useful for anomaly detection in trading platforms.

12 August 2020 in EGUsphere
Verified
Preprint
Open access
PDF file
RCRunlong Cai
Runlong Cai
YLYihao Li
Yihao Li
YCYohann CLEMENT
Yohann CLEMENT
DLDandan Li
Dandan Li
CDClément Dubois
Clément Dubois
MRMatthieu Riva
Matthieu Riva
The Orbitrap mass spectrometer has recently been proved to be a powerful instrument to accurately measure gas-phase and particle-phase organic compounds with a greater mass resolving power than other widely-used online mass spectrometers in atmospheric sciences. We develop an open-source software tool (Orbitool, https://orbitrap.catalyse.cnrs.fr) to facilitate the analysis of long-term online Orbitrap data. Orbitool can average long-term data while maintaining the mass accuracy by re-calibrating each mass spectrum, identify chemical compositions and isotopes of measured signals, and export time series and mass defect plots. The noise reduction procedure in Orbitool can separate signal peaks from noise and greatly reduce the computational and storage expenses. Chemical-ionization Orbitrap data from laboratory experiments on ozonolysis of monoterpenes and ambient measurements in urban Shanghai were used to successfully test Orbitool. For the test dataset, the average mass accuracy was improved from