Dec

Last Updated: February 10, 2023, 11:09 ISTDigiYatra Service (Photo: DigiYatra)The Digi Yatra policy is an initiative launched by the Ministry of Civil Aviation for a biometric boarding system using Facial Recognition TechnologyMore than 1.25 lakh passengers have availed Digi Yatra at three airports since its launch. Digi Yatra was launched at three airports, including Delhi, Bengaluru and Varanasi on December 1, 2022.The Digi Yatra Central Ecosystem (DYCE) is built on the fundamental tenets of privacy by design/default…

New Delhi, February 2 Meta has said that it took down over 22.54 million pieces of content across 13 policies for Facebook and over 12.03 million pieces of content across 12 policies for Instagram in December 2022 in India. Between December 1-31, Facebook received 764 reports through the Indian grievance mechanism, and the company said it provided tools for users to resolve their issues in 345 cases. These include pre-established channels to report content for specific…

How GANs work and how you can use them to synthesize dataFig. 1 — Synthetic images of a person, generated entirely by a GAN. Image source: https://thispersondoesnotexist.com/ . License: https://github.com/lucidrains/stylegan2-pytorch/blob/master/LICENSE (MIT)If you’re working in deep learning, you’ve probably heard of GANs, or Generative Adversarial Networks (Goodfellow et al, 2014). In this post we will explain what GANs are, and discuss some use cases with real examples. I am adding to this post a link to my GAN…

Statistics in R SeriesPhoto by Joshua Sortino on UnsplashIntroductionWe have previously executed a generalized ordinal logistic regression model where the effect of explanatory variables was allowed to vary across different levels of the response variable. When we find out that some of the predictor variables violating the proportional odd assumption, we can allow the effect of only those variables to vary across different levels of outcomes. This type of model is called the partial proportional odd (PPO) model.Background…

Opinion: Evolving AIWhat is the origin of new ideas?What is the origin of new ideas? (Image from Stable Diffusion)The human brain is currently the best Neural Network in the world. Have you ever woken up and felt puzzled about where you were only to realize that you are not in your own home but at someone else’s house where you had stayed the night before? What is the cause of this feeling of disorientation?This is because your brain has learned, over the years, what your room and the world generally look like when you…

A guide to avoiding the common pitfalls of event studiesPhoto by Ricardo Gomez Angel on UnsplashEvent studies are useful tools in the context of causal inference. They are used in quasi-experimental situations. In these situations, the treatment is not randomly assigned. Thus, in contrast to randomized experiments (i.e., A/B tests), one cannot rely on a simple comparison of the means between groups to make causal inferences. In these types of situations, event studies are very useful.Event studies are also frequently used…

A simpler path to Git and understanding main conceptsPhoto by Praveen Thirumurugan on Unsplash“What are all these commands?” “How many hours will it take me to learn them all?” “In what order should I use them?”If those questions sound familiar, you’ve come to the right place.Git can be very frustrating and difficult to grasp, and the amount of information available on the Internet makes it even more overwhelming.We all went through this pain of dealing with Git and being plunged into a black hole from which we can’t…

Learn and improve NLP skills by building your own encoder from scratchA transformer thinking about an essay. Generated by DALL·E 2.There is currently a lot of hype surrounding artificial intelligence (AI) thanks to technologies like ChatGPT. The Transformers, which serve as the fundamental building blocks for many recent popular AI applications, are at the center of this hype. That is why I am writing this article about them. Transformers are deep learning models that have excelled at a variety of natural language…

A way to better understand learning with deep neural networksPhoto by Giulia May on UnsplashCurrently, the theoretical mechanisms of learning with Deep Neural Network (DNN) are not completely well known. One remarkable contribution is the concept of Information Bottleneck (IB) presented by Naftali Tishby in 2017, who was a computer scientist and neuroscientist from the Hebrew University of Jerusalem. His theory claims to be a way to understand the impressive success of neural networks in a huge variety of applications.…

A practitioner’s guide, with a demonstration using the IBM Telco Churn datasetPhoto by Pablo García Saldaña on UnsplashLogistic Regression is commonly used for modeling classification problems. It’s a parametric algorithm whose output provides for powerful model explanations (termed by many as explainable ML). In particular, in addition to overcoming the known limitations of a Linear Regression for modelling classification problems, and in comparison to the non-parametric Tree-based algorithms, it’s able to comfortably…