lawlite19/machinelearning_python

★ 8,482PythonAudience · researcherComplexity · 2/5Setup · easy

Mindmap

mindmap
  root((repo))
    What it does
      Explain ML algorithms
      Show Python code
      Compare to scikit-learn
    Algorithms
      Linear regression
      Neural networks
      Clustering and PCA
    Tech Stack
      Python
      scikit-learn
      NumPy
    Audience
      ML students
      Algorithm learners

mindmap root((repo)) What it does Explain ML algorithms Show Python code Compare to scikit-learn Algorithms Linear regression Neural networks Clustering and PCA Tech Stack Python scikit-learn NumPy Audience ML students Algorithm learners

Click or tap to explore — scroll the page freely

Things people build with this

USE CASE 1

Study how linear regression and gradient descent work by reading commented Python code alongside the math

USE CASE 2

Learn to implement a neural network from scratch in Python using the backpropagation examples

USE CASE 3

Compare a hand-coded K-Means algorithm to the scikit-learn version to understand what the library does internally

Tech stack

Pythonscikit-learnNumPy

Getting it running

Difficulty · easy Time to first run · 30min

Most explanations are written in Chinese, code samples and math notation are readable without it.

In plain English

This repository contains Python implementations of common machine learning algorithms, written primarily in Chinese. It is structured as a study resource that walks through the mathematics behind each algorithm and then shows the corresponding Python code. For each topic, there is also a version using scikit-learn, a popular Python library that provides ready-built versions of the same algorithms. The algorithms covered include linear regression (predicting a numerical output from input data), logistic regression (classifying inputs into categories), a neural network built using backpropagation (a method for training networks layer by layer), support vector machines (a technique for drawing decision boundaries between categories), K-Means clustering (grouping data points into clusters without pre-assigned labels), PCA or principal component analysis (reducing the number of dimensions in a dataset to simplify it), and anomaly detection (identifying data points that deviate significantly from the expected pattern). For each algorithm, the README presents the cost function (a measure of how wrong the model's predictions are), the gradient descent update rule (the step-by-step procedure for improving the model), and a Python code snippet that implements it. Screenshots of results are also included so a reader can see what the output looks like when each algorithm runs. The audience is students and developers learning how these algorithms work under the surface, rather than those who just want to call a library function. Most explanations are in Chinese, so readers who do not read Chinese will find the code samples and mathematical notation more accessible than the prose. The full README is longer than what was shown.

Copy-paste prompts

Prompt 1

Using this repo as reference, help me implement logistic regression from scratch in Python for a binary classification problem

Prompt 2

Explain the gradient descent update rule in this repo's linear regression code and help me adapt it to my own dataset

Prompt 3

Help me understand this repo's neural network backpropagation code and convert it to use modern PyTorch instead

Open on GitHub → Explain another repo

← lawlite19 on gitmyhub — every repo by this author, as a profile.

Verify against the repo before relying on details.