Published in Towards Data Science·May 30An Introduction to Bayesian InferenceA math + code introduction to Bayesian Inference methods — Markov Chain Monte Carlo and Variational Inference. — In the previous blog post, I gave an introduction to the world of Bayesian Statistics. It examined how Bayesian statistics presents a way to incorporate evidence into our model to obtain a better model. This process is known as inference, which is fancy mathematical speak for obtaining/approximating the posterior.Bayesian Statistics8 min read
Published in Towards Data Science·Apr 5Designing a Chemical Reactor with Data ScienceFinal Year Project. The last obstacle between a college student and working life. As a Chemical Engineering student, my final project involved me having to design and optimize a chemical reactor. Personally, my goal was not to get the best grade but approach it as a data science problem, showcasing…Reactor Project7 min read
Published in Towards Data Science·Jan 14An Introductory Primer to Bayesian StatisticsA mathematical introduction to the world of Bayesian Statistics — In 1748, philosopher David Hume wrote the essay “On Miracles” arguing that incredulous and miraculous stories written in the past should not be believed because they went against the law of nature. 15 years later, a posthumous paper by Reverend Thomas Bayes raised an even more important question, “How much…Data Science8 min read
Published in Towards Data Science·Dec 29, 2021Leveling Up Your Machine Learning ProjectsA practical guide on how to structure machine learning projects — Starting out in data science, notebooks are your friend. They are great at being multi-purpose tools to visualize and explore the data but are not the best as your project becomes more complicated. For such projects, we will want code that is more structured and scalable to work with different…Data Science8 min read
Published in Towards Data Science·Dec 14, 20215 Essential Concepts for New Data ScientistsData science can be extremely intimidating especially when just starting out. While it can be easy to use plug-and-play code snippets implementing popular machine learning algorithms, it cannot substitute for mathematical intuition. Here are some important concepts which I found to be rather essential to my data science journey. 1. Bias Variance Trade-off The…Data Science9 min read
Published in Towards Data Science·Sep 18, 2021Understanding RegularisationRegularisation is a powerful tool that can help tune models — At the start of this year, I had a very limited understanding of regularisation. To me, it was simply a tool to use when overfitting. If the model was having higher validation loss than train loss, adding L1 or L2 regularisation might be able to solve this issue. …Data Science9 min read
Published in Towards Data Science·Jul 2, 2021Building a VAE Playground with StreamlitWith Streamlit creating a deploying a web app can be really easy! — This blog post is part of a mini-series that talks about the different aspects of building a PyTorch Deep Learning project using Variational Autoencoders. Part 1: Mathematical Foundations and Implementation Part 2: Supercharge with PyTorch Lightning Part 3: Convolutional VAE, Inheritance and Unit Testing Part 4: Deployment In previous sections…Data Science8 min read
Published in Geek Culture·Jun 21, 2021My Data Science Internship: The Start-up experienceTwo weeks ago, I ended a 6 months internship at AiPalette. I have learnt so much from working there and have walked away a much better data scientist. I learnt a lot more from this than I initially expected. While big tech companies are still the dream job for most…Artificial Intelligence8 min read
Published in Towards Data Science·Jun 10, 2021Beginner Guide to Variational Autoencoders (VAE) with PyTorch Lightning (Part 3)Building a new model with minimal additional code — This blog post is part of a mini-series that talks about the different aspects of building a PyTorch Deep Learning project using Variational Autoencoders. Part 1: Mathematical Foundations and Implementation Part 2: Supercharge with PyTorch Lightning Part 3: Convolutional VAE, Inheritance and Unit Testing Part 4: Streamlit Web App and…Pytorch8 min read
Published in Towards Data Science·May 3, 2021Beginner guide to Variational Autoencoders (VAE) with PyTorch Lightning (Part 2)This blog post is part of a mini-series that talks about the different aspects of building a PyTorch Deep Learning project using Variational Autoencoders. Part 1: Mathematical Foundations and Implementation Part 2: Supercharge with PyTorch Lightning Part 3: Convolutional VAE, Inheritance and Unit Testing Part 4: Streamlit Web App and…Machine Learning6 min read
