Summary of Explaining Model Overfitting in Cnns Via Gmm Clustering, by Hui Dou et al.
Explaining Model Overfitting in CNNs via GMM Clustering
by Hui Dou, Xinyu Mu, Mengjun Yi, Feng Han, Jian Zhao, Furao Shen
First submitted to arxiv on: 12 Dec 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
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 This paper investigates the decision-making processes of Convolutional Neural Networks (CNNs) in computer vision tasks. While CNNs have achieved impressive results, their lack of transparency hinders practical applications. The authors propose a novel approach to assess CNN filters by clustering feature maps using Gaussian Mixture Model (GMM). By analyzing these clusters, they identify anomaly filters associated with outlier samples and explore the relationship between these filters and model overfitting. This method is universally applicable across various CNN architectures, including AlexNet and LeNet-5, as demonstrated through three meticulously designed experiments. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper tries to make Convolutional Neural Networks (CNNs) more understandable. Right now, it’s hard to figure out how they make decisions. The authors came up with a new way to look at the filters inside CNNs, called Gaussian Mixture Model (GMM). By using this method, they can identify weird filters that are connected to strange data points. They also explored why these weird filters might be causing the model to get worse and worse over time. This idea works for all kinds of CNN models and can help us understand how they work better. |
Keywords
» Artificial intelligence » Clustering » Cnn » Mixture model » Overfitting
