river
river copied to clipboard
online-ml
🌊 Online machine learning in Python
I took a first stab at metric learning but I'm having trouble making it work. [Here](https://github.com/creme-ml/creme/files/3639804/Metric.learning.zip) is a notebook that contains what I tried. Here are some references: - [Online...
- [https://www.unofficialgoogledatascience.com/2021/04/why-model-calibration-matters-and-how.html](https://www.unofficialgoogledatascience.com/2021/04/why-model-calibration-matters-and-how.html)
[drunk-main.zip](https://s3-us-west-2.amazonaws.com/secure.notion-static.com/fc2642bc-dba1-463d-a59e-15e1cdb3a12f/drunk-main.zip) *Implement it in-memory, then figure out out-of-core details.* ### **Resources** - [DeepWalk: Online Learning of Social Representations](https://arxiv.org/pdf/1403.6652.pdf) 🖨 - [Why I'm lukewarm on graph neural networks](https://www.reddit.com/r/MachineLearning/comments/kqazpd/d_why_im_lukewarm_on_graph_neural_networks/) - [700x faster...
[https://amitness.com/2020/08/information-retrieval-evaluation/](https://amitness.com/2020/08/information-retrieval-evaluation/)
- [One Pass ImageNet](https://arxiv.org/abs/2111.01956) - [Prioritized Experience Replay](https://arxiv.org/abs/1511.05952)
- [https://gist.github.com/agramfort/2071994](https://gist.github.com/agramfort/2071994) - [https://medium.com/filament-ai/making-text-search-learn-from-feedback-4fe210fd87b0](https://medium.com/filament-ai/making-text-search-learn-from-feedback-4fe210fd87b0) - [https://www.wikiwand.com/en/Ordinal_regression](https://www.wikiwand.com/en/Ordinal_regression) [Learning to rank with scikit-learn: the pairwise transform](https://fa.bianp.net/blog/2012/learning-to-rank-with-scikit-learn-the-pairwise-transform/)
[Suggestion: no more return self · Discussion #935 · online-ml/river](https://github.com/online-ml/river/discussions/935)
[Professional-grade mypy configuration - Wolt Blog](https://blog.wolt.com/engineering/2021/09/30/professional-grade-mypy-configuration/)
- [pandas-stubs](https://github.com/pandas-dev/pandas-stubs)
[ta-lib](https://ta-lib.org/) is a famous library for doing technical analysis, which analyses financial data. There is a Python binding [here](https://mrjbq7.github.io/ta-lib/), with a [streaming API](https://github.com/mrjbq7/ta-lib#streaming-api). It would be great to write an...