Simulate a tiny solar system with Sun, Earth, Mars, and an unknown comet using real mass, distance, and velocity data in Matplotlib — I was amazed by the PCA (Principal Components Analysis) and thinking of using animation to show the machine learning processes. When I managed to animate a graph, I was tempted to animate something cool. The first cool stuff that comes to me is the solar system. This article is also…

Understand the K-Means model by creating one from scratch — K-Means model is one of the unsupervised machine learning models. This model is usually used to partition observed data into k clusters. You give the model a bunch of data with defined features and tell it how many clusters you want it to output. …

Walk around amazon.com's anti-bot mechanism and use Python to crawl Amazon.com price data automatically — It is not that easy to crawl Amazon.com data nowadays. Amazon will block your HTTP requests here and there, and make your automation a vain trial. For example, run the following code: import requests page = requests.get('https://www.amazon.com/') print(page.text) Your request will get a warm welcome from Amazon.com like this instead…

Build a neural network model to simulate any functions using PyTorch — We know that the Linear Regression model can help generate a line to simulate “Linear Data”. A function in the form of f(x) = ax + b. So, what about other functions, like non-linear ones? Even an irregular shape that we know nothing about. Can we still simulate a model…

Leverage the idea from Neural Network to estimate weights for a system with multiple factors — Imagine you have a customer dataset with dozens of features, and you are trying to build a weighted system to predict certain customer behaviors (like buy or not buy). How to determine those individual weights value?

Understand how image Fourier Transform works and implement it with the shortest Python code — Image 2D Fourier Transform is so useful and powerful, that you will use it sooner or later if you work with Machine Learning or vision automation. Call API or libraries to use it is a piece of cake, fully understanding 2D Discrete Fourier Transform isn’t that easy. In this article…

Understand convolution with a real case and verify the result with Python — Convolution is an operation commonly used in digital signal processing and neural network models. Most DSP or ML textbook explains the Convolution operation with lots of math equations and complex mathematic derivations. Looks scary and easily gets confused. In this article, I am trying to explain the operation with a…

An Anatomy of Arithmetic Mean and Geometric Mean — Two Cases A stock has half the chance to increase 30% monthly and another half to decrease 25% monthly. I will say, hm, it is a good stock. Averagely speaking, every month I will have a 5% profit if I keep holding it for a year. I will get 79% (1.05¹² =…

An easy way to create and manage VMs in Linux, Windows and macOS with the same commands in pure terminal — WSL/WSL2 is a phenomenal tool in Windows, it just so much empowers Windows users to use Linux, Sometimes I feel Windows is the best desktop to run Linux. But what about Linux/Ubuntu and macOS, is there anything similar to WSL of Windows? Then I found Multipass from canonical, the same…