Data Science at Home

In the second episode of Rust and Machine learning I am speaking with Luca Palmieri, who has been spending a large part of his career at the interception of machine learning and data engineering. In addition, Luca contributed to several projects closer to the machine learning community using the Rust programming language. Linfa is an ambitious project that definitely deserves the attention of the data science community (and it's written in Rust, with Python bindings! How cool??!).   References Series Announcement - Zero to Production in Rust https://www.lpalmieri.com/posts/2020-05-10-announcement-zero-to-production-in-rust/ Zero To Production #0: Foreword https://www.lpalmieri.com/posts/2020-05-24-zero-to-production-0-foreword/ Taking ML to production with Rust: a 25x speedup https://www.lpalmieri.com/posts/2019-12-01-taking-ml-to-production-with-rust-a-25x-speedup/

This is the first episode of a series about the Rust programming language and the role it can play in the machine learning field. Rust is one of the most beautiful languages I have ever studied so far. I personally come from the C programming language, though for professional activities in machine learning I had to switch to the loved and hated Python language. This episode is clearly not providing you with an exhaustive list of the benefits of Rust, nor its capabilities. For this you can check the references and start getting familiar with what I think it's going to be the language of the next 20 years.   Sponsored This episode is supported by Pryml Technologies. Pryml offers secure and cost effective data privacy solutions for your organisation. It generates a synthetic alternative without disclosing you confidential data.   References The Rust Programming Language  Cookin' with Rust  

In this episode I have a chat with Sandeep Pandya, CEO at Everguard.ai a company that uses sensor fusion, computer vision and more to provide safer working environments to workers in heavy industry. Sandeep is a senior executive who can hide the complexity of the topic with great talent.   This episode is supported by Pryml.io Pryml is an enterprise-scale platform to synthesise data and deploy applications built on that data back to a production environment. Test ideas. Launch new products. Fast. Secure.

As a continuation of the previous episode in this one I cover the topic about compressing deep learning models and explain another simple yet fantastic approach that can lead to much smaller models that still perform as good as the original one. Don't forget to join our Slack channel and discuss previous episodes or propose new ones. This episode is supported by Pryml.io Pryml is an enterprise-scale platform to synthesise data and deploy applications built on that data back to a production environment.   References Comparing Rewinding and Fine-tuning in Neural Network Pruning https://arxiv.org/abs/2003.02389  

Using large deep learning models on limited hardware or edge devices is definitely prohibitive. There are methods to compress large models by orders of magnitude and maintain similar accuracy during inference. In this episode I explain one of the first methods: knowledge distillation  Come join us on Slack Reference Distilling the Knowledge in a Neural Network https://arxiv.org/abs/1503.02531 Knowledge Distillation and Student-Teacher Learning for Visual Intelligence: A Review and New Outlooks https://arxiv.org/abs/2004.05937

Codiv-19 is an emergency. True. Let's just not prepare for another emergency about privacy violation when this one is over.   Join our new Slack channel   This episode is supported by Proton. You can check them out at protonmail.com or protonvpn.com

Whenever people reason about probability of events, they have the tendency to consider average values between two extremes.  In this episode I explain why such a way of approximating is wrong and dangerous, with a numerical example. We are moving our community to Slack. See you there!    

In this episode I briefly explain the concept behind activation functions in deep learning. One of the most widely used activation function is the rectified linear unit (ReLU).  While there are several flavors of ReLU in the literature, in this episode I speak about a very interesting approach that keeps computational complexity low while improving performance quite consistently. This episode is supported by pryml.io. At pryml we let companies share confidential data. Visit our website. Don't forget to join us on discord channel to propose new episode or discuss the previous ones.  References Dynamic ReLU https://arxiv.org/abs/2003.10027

One of the best features of neural networks and machine learning models is to memorize patterns from training data and apply those to unseen observations. That's where the magic is.  However, there are scenarios in which the same machine learning models learn patterns so well such that they can disclose some of the data they have been trained on. This phenomenon goes under the name of unintended memorization and it is extremely dangerous. Think about a language generator that discloses the passwords, the credit card numbers and the social security numbers of the records it has been trained on. Or more generally, think about a synthetic data generator that can disclose the training data it is trying to protect.  In this episode I explain why unintended memorization is a real problem in machine learning. Except for differentially private training there is no other way to mitigate such a problem in realistic conditions. At Pryml we are very aware of this. Which is why we have been developing a synthetic data generation technology that is not affected by such an issue.   This episode is supported by Harmonizely.  Harmonizely lets you build your own unique scheduling page based on your availability so you can start scheduling meetings in just a couple minutes. Get started by connecting your online calendar and configuring your meeting preferences. Then, start sharing your scheduling page with your invitees!   References The Secret Sharer: Evaluating and Testing Unintended Memorization in Neural Networks https://www.usenix.org/conference/usenixsecurity19/presentation/carlini

In this episode I explain a very effective technique that allows one to infer the membership of any record at hand to the (private) training dataset used to train the target model. The effectiveness of such technique is due to the fact that it works on black-box models of which there is no access to the data used for training, nor model parameters and hyperparameters. Such a scenario is very realistic and typical of machine learning as a service APIs.  This episode is supported by pryml.io, a platform I am personally working on that enables data sharing without giving up confidentiality.    As promised below is the schema of the attack explained in the episode.     References Membership Inference Attacks Against Machine Learning Models