wepe/machinelearning

★ 5,688PythonAudience · researcherComplexity · 2/5Setup · easy

Mindmap

mindmap
  root((wepe/ml))
    Deep Learning
      CNN implementations
      MLP networks
      Softmax regression
      MNIST digit recognition
    Traditional ML
      KNN and SVM
      Decision trees
      Naive Bayes
      PCA and clustering
    Frameworks
      Keras
      Theano
      Python
    Datasets
      MNIST digits
      Face recognition
    Learning Resources
      CSDN blog posts
      Theory walkthroughs
    Applications
      Image classification
      Data visualization

mindmap root((wepe/ml)) Deep Learning CNN implementations MLP networks Softmax regression MNIST digit recognition Traditional ML KNN and SVM Decision trees Naive Bayes PCA and clustering Frameworks Keras Theano Python Datasets MNIST digits Face recognition Learning Resources CSDN blog posts Theory walkthroughs Applications Image classification Data visualization

Click or tap to explore — scroll the page freely

Things people build with this

USE CASE 1

Study classic ML algorithm implementations from scratch alongside detailed Chinese-language blog explanations.

USE CASE 2

Experiment with CNN and MLP architectures for handwritten digit recognition using the MNIST dataset.

USE CASE 3

Learn traditional ML methods like SVM, k-means, PCA, and Naive Bayes through working Python code.

USE CASE 4

Use as a reference collection when learning how deep learning frameworks like Keras and Theano work.

Tech stack

PythonKerasTheanoNumPy

Getting it running

Difficulty · easy Time to first run · 30min

No installation or package required. Clone the repo, install Python with Keras and Theano, and run individual scripts. Reading Chinese is helpful for the accompanying CSDN blog posts.

No license is mentioned in the explanation.

In plain English

This repository is a personal study collection of machine learning and deep learning algorithm implementations, written mostly in Python. The content is in Chinese. The author compiled it while learning the subject and notes that mistakes may exist, since they were a beginner at the time. The collection covers a range of classic algorithms. On the deep learning side there are implementations of convolutional neural networks (CNNs), multi-layer perceptrons (MLPs), and softmax regression, mostly using the Keras and Theano frameworks. These include applications like handwritten digit recognition using the MNIST dataset and a face recognition demo. On the traditional machine learning side, the repository includes implementations of k-nearest neighbors, logistic regression, support vector machines (SVM), decision trees, k-means clustering, Naive Bayes classifiers, Gaussian Mixture Models, principal component analysis (PCA), ridge regression, and manifold learning for data visualization. Each section typically links to a blog post (on the Chinese platform CSDN) where the author explains the theory and walks through the code in detail. The code itself serves as the working implementation alongside those write-ups. This is a learning resource rather than a production library. There is no installation process and no packaged API. It is best suited for people who want to study algorithm implementations from scratch alongside explanatory articles, and who can read Chinese or are comfortable navigating the code directly.

Copy-paste prompts

Prompt 1

Explain how the SVM implementation in this repo works and how it differs from sklearn's SVC. Show me the key math steps in the code.

Prompt 2

Walk me through the CNN architecture used for MNIST digit recognition in this repo. What layers are used and why?

Prompt 3

Compare the PCA implementation here to a standard textbook version. Are there any simplifications or mistakes a beginner might make?

Prompt 4

How does the k-means clustering code in this repo handle initialization and convergence? Can you suggest improvements?

Prompt 5

I want to adapt the Naive Bayes classifier from this repo to classify English text. What changes do I need to make?

Open on GitHub → Explain another repo

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

Verify against the repo before relying on details.