To all organizers, developers, and maintainers involved in the Scikit-learn project,
I would like to share a recent article that researchers from MIT, ETH, and Kyoto University (myself) have published about building efficient models for drug discovery and pharmaceutical data mining. In short, it demonstrates through replicate experiment that neither big data nor complex AI such as deep learning are necessary for efficient drug discovery, and that active learning can guide/assist decision making processes in the real world. The paper's success is underpinned by the use of Scikit-learn's RandomForestClassifier implementation combined with other techniques developed in the work. Therefore, it is a by-product of the volunteerism, hard work, and dedication by those involved in scikit-learn. As the senior author of this study, I wish to share my great appreciation for your efforts. While I am strongly limited in time and can barely contribute to this community, I cannot thank all of you enough for your work - it has made an impact. We are working on theoretical extensions of the work now, as well as pushing the technology forward in applied discovery sciences (in agricultural, pharmaceutical, and medical areas). In the theory and real-world applications, scikit-learn is indispensible. We have made the paper open access, and hope that such will inspire this community as well as those in applied sciences. You will see that the open source software community has been listed in the Acknowledgments. Certainly, we would welcome even the most casual of comments about the paper. The paper can be retrieved from here: http://www.future-science.com/doi/abs/10.4155/fmc-2016-0197 With kindest regards and sincere appreciation, J.B. Brown Kyoto University Graduate School of Medicine Junior Associate Professor and Principal Investigator http://statlsi.med.kyoto-u.ac.jp/~jbbrown PS - To those of you involved in the matplotlib, scipy, and numpy projects, your forwarding of this to those projects would be appreciated. They were also critical.
_______________________________________________ scikit-learn mailing list [email protected] https://mail.python.org/mailman/listinfo/scikit-learn
