Timeline for Can you use neural networks with limited data?
Current License: CC BY-SA 4.0
12 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Feb 10, 2020 at 9:14 | history | edited | Adam | CC BY-SA 4.0 |
typo
|
| Feb 9, 2020 at 10:34 | comment | added | Paze | Thank you (there is a bracket in your link that returns a 404 though, but I got it). | |
| Feb 8, 2020 at 22:22 | comment | added | Adam | Machine learning is typically described as involving one of several different types of learning, such as supervised, semisupervised, unsupervised, and reinforcement. NNs are one possible type of machine learning model to consider for these problems [for example, see the nice flowchart at kdnuggets.com/2017/06/which-machine-learning-algorithm.html] | |
| Feb 8, 2020 at 22:08 | comment | added | Paze | That's interesting I thought NN/deep learning and machine learning were two different subjects. | |
| Feb 8, 2020 at 22:04 | comment | added | Adam | @Paze, two things. First, yes to your general question, NN is not necessarily better, but worth trying in many situations. Second, just to clarify (helping for future questions), people typically refer to NN as a form of machine learning. I would word your phrase as, "NN is not necessarily better than other machine learning algorithms/models." | |
| Feb 8, 2020 at 21:47 | comment | added | Paze | Thank you, that answers my question. NN is not necessarily better than machine learning for smaller N but may be, we just have to find out via comparison. Is that correctly understood? | |
| Feb 8, 2020 at 21:47 | vote | accept | Paze | ||
| Feb 8, 2020 at 21:45 | comment | added | Wayne | +1. If you have small data, a Bayesian approach makes the most sense of all, since you can inject information into your model via priors. | |
| Feb 8, 2020 at 21:38 | comment | added | Adam | @Paze, I've updated the answer with comparison suggestions. | |
| Feb 8, 2020 at 21:37 | history | edited | Adam | CC BY-SA 4.0 |
added content answering question in comment
|
| Feb 8, 2020 at 21:14 | comment | added | Paze | Hi Adam I'm very interested in hearing how you compared the NN results to your regression results as this would be my next step? | |
| Feb 8, 2020 at 21:09 | history | answered | Adam | CC BY-SA 4.0 |