Summary of Deep Neural Networks For Choice Analysis: Enhancing Behavioral Regularity with Gradient Regularization, by Siqi Feng et al.
Deep neural networks for choice analysis: Enhancing behavioral regularity with gradient regularization
by Siqi Feng, Rui Yao, Stephane Hess, Ricardo A. Daziano, Timothy Brathwaite, Joan Walker, Shenhao Wang
First submitted to arxiv on: 23 Apr 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
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 proposes novel metrics, strong and weak behavioral regularities, to evaluate the monotonicity of individual demand functions in deep neural networks (DNNs) for travel behavior modeling. The study designs a constrained optimization framework with six gradient regularizers to enhance DNNs’ behavioral regularity. Applying this framework to Chicago and London travel survey data, the results show that benchmark DNNs cannot guarantee behavioral regularity but gradient regularization (GR) increases it by around 6 percentage points while retaining predictive power. GR is more effective in small sample scenarios, improving behavioral regularity by about 20 percentage points and log-likelihood by around 1.7%. The paper highlights the importance of behavioral regularization for enhancing model transferability and application in forecasting. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This research looks at how well a type of artificial intelligence called deep neural networks (DNNs) can understand human behavior when it comes to travel. Currently, DNNs don’t always make sense or follow rules, which limits their usefulness. The researchers created new ways to measure how regular or predictable DNNs are in understanding travel behavior. They then used these measurements on real data from Chicago and London. Their results showed that using special techniques, called gradient regularization, can improve the predictive power of DNNs while making them more regular and easier to understand. |
Keywords
» Artificial intelligence » Log likelihood » Optimization » Regularization » Transferability
