cbhihe / smartcity_high-dim-statistical-learning Goto Github PK

Some of the work presented here is concerned with mere toy examples meant for others to hone their ML skills. They are mostly those in folders named Lab1, Lab2, etc. Other developments are more personal and involve a greater amount of coding using R and / or python scripting. This repo also contains a project in deep learning, housed in the folder Project, with its own introductory readme.md file.
Altogether various areas are covered by the contents in a more or less detailed fashion:

• regression (linear and beyond)
• classification (incl. generative and discriminative classifiers)
• clustering: k-means and E-M
• learning with kernels: SVM and kernel functions
• artificial neural networks (NN): Delta rule, MLP-1, MLP-2, RBF, DL and CNNs
• ensemble methods: Random Forests

Some of the public domain data set used to perfom analyses were not uploaded on this repo due to the size limitation on individual files imposed by GitHub. In such cases at least a snapshot of the data sets is provided as well as a URL, where it can be downloaded. The downloaded data set may or may not be identical to the one originally used to obtain results presented here. The responsability for that falls squarely on the data maintainer. In practice however, interested people may still benefit from the strategy laid out to deal with the data, irrespectively of whether some of the data has changed. Data set changes over the years (if any) are likely to be all or nothing in any case; i.e. a data set may altogether cease to be available because the website has disappeared or moved.

Feel free to copy any material found here and to further release it to the general public, subject to the terms of the GNU General Public License v3... That license is not a right granted to you to plagiarize the content of the repo. Your editor, publisher, university, even your mother and something called common decency all have rules against that. So do courts of law in most countries. So use common sense and live happily ever after.

TLDR: Basically the material on this repo is available for free (as in "free beer") as long as you always attach the above mentioned license's terms to it and you do not include it in other work, program, product or applications (commercial in nature or not) which do not comply with the GNU General Public License v3 or with something "very very very close". Of course the above one sentence summary cannot be considered authoritative in terms of licensing. It does not replace the GNU General Public License v3 included in this repo, which recoups the complete licensing terms for your reading pleasure. ;-)

The copyright however stays with me and with any contributors to this repo, according to who contributed what to the contents of the repo.