Summary of Usage-specific Survival Modeling Based on Operational Data and Neural Networks, by Olov Holmer et al.
Usage-Specific Survival Modeling Based on Operational Data and Neural Networks
by Olov Holmer, Mattias Krysander, Erik Frisk
First submitted to arxiv on: 27 Mar 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Systems and Control (eess.SY); Machine Learning (stat.ML)
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 A novel approach for modeling component failure times using neural network-based survival models is presented, which can be particularly useful in planning maintenance tasks. Conventional survival models are trained using snapshot data gathered and stored at specific times, but this data often lacks independence due to multiple snapshots from the same individual. The authors propose a method for resampling non-homogeneously sampled data to make it homogeneously sampled, allowing for maximum likelihood training. They also introduce a technique for randomly resampling the dataset during each epoch of training to reduce its size and improve training efficiency. The proposed methodology is evaluated on both simulated and experimental datasets of starter battery failures, demonstrating its effectiveness in producing accurate survival models. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary A team of researchers developed a new way to predict when things will break down, which can be really helpful for maintenance work. They used special kinds of computer models called neural networks that take into account the timing of when we know something has failed before. The problem is that this data isn’t always independent, meaning it’s not just separate bits of information, but often comes from the same thing that failed more than once. To fix this, they came up with a way to reorganize the data so that it’s easier to work with. They also found a way to make the training process faster by randomly sampling the data at each step. |
Keywords
* Artificial intelligence * Likelihood * Neural network
