[ATOM] Advanced Topics on Machine Learning discussion group

ATOM is meeting on Tuesday, 23rd October, 6:30pm, at Galvanize!

This month, Susan Fung will lead a discussion about attacks on training data.

https://arxiv.org/pdf/1804.07933.pdf
Article title = Is feature selection secure against training data poisoning?

The existence of the internet has led to the invention of a lot of great services, however, there are plenty of agents out there inventing ways to maliciously profit from an individual’s or company’s information. Machine learning algorithms utilized by engineers and data scientists are valuable tools in data security, but the interpretation of the results relies on the quality of the data, and the features feeding into the algorithm. When the feature selection process is targeted by the malicious agent, this is known as a “poisoning attack”. The goal of this paper is to propose a way to categorize poisoning attacks which means understanding different ways feature selection can be attacked. The authors start from what is known about defending against attacks on supervised and unsupervised algorithms, and present attack strategies using LASSO, Ridge, and elastic net - popular methods used for feature selection.

About ATOM:
Advanced Topics on Machine learning ( ATOM ) is a learning and discussion group for cutting-edge machine learning techniques in the real world. Read our mission statement here : https://gist.github.com/QCaudron/3c405569acebbf75b6bd02fd1bb17480

As a discussion group, we strongly encourage participation, so be sure to read up about the topic of conversation beforehand !

ATOM can be found on PuPPy’s Slack under the channel #atom, and on PuPPy’s Meetup.com events.

We're kindly hosted by Galvanize (https://www.galvanize.com). Thank you !