Published in Towards Data Science·PinnedMember-onlyLatent Variables & Expectation Maximization AlgorithmBayesian Approach to Machine Learning — ‘Latent’, originated from Latin which means being hidden. Probably you have all known about latent heat, which is the heat energy required for phase transformation keeping the temperature constant. So we observe a change, but the reason behind it is apparently hidden. The motivation for Latent Variable Model (LVM) is…Machine Learning14 min readMachine Learning14 min read
Published in Towards Data Science·Mar 8Member-onlyNew Scikit-Learn is More Suitable for Data AnalysisPandas Compatibility and More in Scikit-Learn Version ≥1.2.0 — Around last year December, Scikit-Learn released a major stable update (v. 1.2.0–1) and finally I get to try some of the highlighted new features. It’s now more compatible with Pandas and a few other features will also help us in regression as well as classification tasks. Below, I go through…Data Science5 min readData Science5 min read
Published in Towards Data Science·Feb 21Member-onlyUnderstand & Implement Masked AutoRegressive Flow with TensorFlowFlow-Based Models for Density Estimation with TensorFlow — Previously we went through the details of the mathematics behind Normalizing Flows and some examples of transforming probability distributions. Here we combine all these concepts to understand Autoregressive Flows and how to implement them using TensorFlow Probability library. What can you expect from this post — Why Triangular Matrices are…Deep Learning8 min readDeep Learning8 min read
Published in A Bit of Qubit·Jan 26Member-onlyUnderstanding Bloch Sphere: From Density Matrix PerspectivePure & Mixed States, Density Matrix & Bloch Sphere — This post will cover the fundamentals of representing a single qubit state using the Bloch sphere from a mathematical perspective of the density matrix. What will be covered in the post? What is the difference between pure and mixed states? What is a density matrix? How does it help to…Qubit11 min readQubit11 min read
Published in The Startup·Jan 10Member-onlyFuture of Writers in the Era of ChatGPTCreativity and Nuances Matter: According to ChatGPT! — Response from ChatGPT: The following is a generated response by ChatGPT with the search line “Future of writers in the era of chatgpt” >>> As with any technological advancement, chatbots and language models like GPT (Generative Pre-trained Transformer) are likely to have some impact on the field of writing. However, it is important…Writing4 min readWriting4 min read
Published in Towards Data Science·Nov 14, 2022Member-onlyTransforming Probability Distributions & Normalizing FlowsUsing Simple Bijectors to Transform Distributions using TensorFlow — What you can expect to learn/review from this post — Getting started with TransformedDistribution module (within TensorFlow Probability library) to transform a base distribution via a bijection operation. Writing our own custom bijection function in TensorFlow. How to define the forward and inverse transformation within a custom bijection function?Probability5 min readProbability5 min read
Published in Towards Data Science·Sep 27, 2022Member-onlyGetting Started with Normalizing Flows: Linear Algebra & ProbabilityChange of Variables Rule, Bijection & Diffeomorphism — The basis of generative modelling is to understand the distribution from where the data samples came from. In one of my previous posts, an example of generative modelling was demonstrated by going through the steps of the Expectation-Maximization algorithm, where we assumed that latent variables give rise to the expected…Deep Learning8 min readDeep Learning8 min read
Published in Towards Data Science·Jul 13, 2022Member-onlyBayesian Deep Learning & Estimating UncertaintyWeather Data; Aleatoric & Epistemic Uncertainty — Performances of Deep Neural Networks (DNNs) rely on the ability to progressively build and extract features from large data. Though these deep models are usually adaptive the performance depends on the data distribution. The robustness property is important in various applications such as computer vision tasks, eg. autonomous driving, because…Deep Learning9 min readDeep Learning9 min read
Published in A Bit of Qubit·Jun 17, 2022Member-onlyShor’s Algorithm: How Does it Work?Finding Prime Factors of Integer using Quantum Computer — We all have possibly read stories and news about the encryption schemes to protect our credit cards, confidential files will be broken by the advancement of quantum computing technology in recent years. Here’s another recent Nature article (from this year, February) about it and how we can safeguard our privacy…Quantum Computing8 min readQuantum Computing8 min read
Published in Towards Data Science·Feb 9, 2022Member-onlyUnderstand and Implement Vision Transformer with TensorFlow 2.0Self-Attention Mechanism and Goodbye Convolution! — When Transformer Network came out, initially it became the go to model for NLP tasks. ‘An Image is Worth 16X16 Words’ which was presented in International Conference for Representation Learning (ICLR) 2021, by Alex Dosovitskiy et.al. showed for the first time how Transformer can be implemented for Computer Vision tasks…Artificial Intelligence13 min readArtificial Intelligence13 min read
