Published inGeek Culture·Apr 1, 2021How Visualizations Help Make Sense of Deep LearningDeep learning poses many challenging questions. Visualization, through its exploratory and explanatory powers, can shed light on some of them: Why — why does the model work (interpretability) How — how well might the model work (on different subsets of data) What — what’s the performance of the model We’ll…Data Visualization6 min readData Visualization6 min read
Feb 4, 2021One month of generative art explorationsAt the beginning of 2021, I participated in this creative coding event called genuary. Each day one receives a prompt, such as ‘do not repeat’ or ‘no loop’. …Creative Coding5 min readCreative Coding5 min read
Jun 16, 2020The unreasonable flexibility of flow fields in generative artVector-based creative coding 🎨 — Flow fields are basically a grid of vectors. One can create all kinds of visuals based on it. A stunning example is the Wind Map by Fernanda Viégas and Martin Wattenberg with a fur-like texture visualizing the speed of the wind. Today I’ll demonstrate how to cook up some unique…Generative Art4 min readGenerative Art4 min read
Published inTowards Data Science·Mar 23, 2020Understand the machine learning Blackbox with ML-interpreterA web app for auto-interpreting the decisions of algorithms like XGBoost — There are dangers in having models running the world and making decisions from hiring to criminal justice. While it’s ideal to have models that are both interpretable & accurate, many of the popular & powerful algorithms are still black-box. Among them are highly performant tree ensemble models such as lightGBM…Machine Learning8 min readMachine Learning8 min read
Oct 8, 2019Takeaways from Strata Data Conference in NYCStrata is a conference focused on data science and enterprise AI/ML. Recently I got a chance to attend and observed some new developments in the field. Here I’ll highlight some talks and concepts I find interesting. Visual & Hand-drawn queries as data input In a talk on data science for…Machine Learning3 min readMachine Learning3 min read
Jan 29, 2019Drawing waves with p5.jsRecently I’ve been playing with p5.js to make generative sketches. p5 is a processing-based library which makes it very flexible to sketch shapes & movements. A particular function that impressed me is perlin noise, which is an Oscar-winning algorithm for generating natural looking random noise. The basic code to draw…JavaScript2 min readJavaScript2 min read
Jun 10, 2018The reigns and ruin of the Roman EmpireToday I looked into the dynasties of the Roman Empire spanning from 27 BC till its split in 395 AD. To encode the different dimensions of the dataset, I used a glow to show that the ruler rose to power via birthright, used the direction to show what happened to…History2 min readHistory2 min read
May 22, 2018Which Sport is the Toughest of Them AllToday I looked into the level of skills required by 60 different sports to find out how tough some sports are. Looking at the distribution of the skills required, we can see most sports tend to require a higher level of endurance rather than analytical aptitude. Narrowing down to the…Sports2 min readSports2 min read
May 13, 2018School shooting in numbersToday I looked into the school shooting incidents from 2010 to 2018 in US primary, middle and high schools with data compiled by the Washington Post. To compare each state in terms of severity and the nature of shooting for the past 2 decades, I used percentage of targeted attack…Data Visualization1 min readData Visualization1 min read
Apr 28, 2018Where does rice come fromToday I looked into world food balance sheet from Food and Agriculture Organization of the United Nations (FAO) to figure out which are the main countries producing, importing and exporting rice. Note that the last update of the database was 2013. While the percentage of big producers might have not…Food2 min readFood2 min read
