Summary of Block-as-domain Adaptation For Workload Prediction From Fnirs Data, by Jiyang Wang et al.
Block-As-Domain Adaptation for Workload Prediction from fNIRS Data
by Jiyang Wang, Ayse Altay, Senem Velipasalar
First submitted to arxiv on: 30 Apr 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Human-Computer Interaction (cs.HC); Signal Processing (eess.SP)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed class-aware-block-aware domain adaptation (CABA-DA) method addresses the challenge of predicting cognitive workload from functional near-infrared spectroscopy (fNIRS) data by minimizing intra-session variance and maximizing inter-class domain discrepancy. This is achieved by viewing different blocks from the same subject in the same session as different domains. The CABA-DA method is combined with an MLPMixer-based model for cognitive load classification, which outperforms three baseline models on three public-available datasets. These datasets include n-back tasks and finger tapping. The proposed contrastive learning method also improves baseline performance. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Predicting how hard someone’s brain is working from a special kind of light measurement is important for understanding how people learn and think. Right now, there are many different ways to do this, but most of them only work well if the training data comes from the same people or sessions as the testing data. This makes it hard to use these methods in real-world situations. The proposed method tries to fix this problem by viewing different blocks of data from the same person on the same day as coming from different “domains”. This helps the model learn to recognize patterns that are unique to each block, rather than just relying on general patterns that don’t change much over time. The proposed method also uses a special type of neural network and a way of learning called contrastive learning to improve performance. |
Keywords
» Artificial intelligence » Classification » Domain adaptation » Neural network
