whuang022ai Goto Github PK

followers: 22.0 following: 7.0 repos: 21.0 gists: 8.0

Name: 楊明翰

Type: User

Bio: Computer vision algorithms +Machine learning + Natural language processing

Location: Taiwan Taipei

Blog: https://1fly2sky.wordpress.com/

楊明翰's Projects

ai-tutorial-multilayer-perceptron

This is a hard coding version of Sigmoid Multilayer Perceptron with 2 input *2 hideen *1 output that can slove XOR problem. It is just for "Hello World" for the A.I beginners.

ai-tutorial-multilayer-perceptron-numpy-batch-version

This is a 3 layer MLP Neural Network using only Numpy that can help us understand the principle of the neural network.This is a version that processes the data by full batch which is more easier than the older version.

ai-tutorial-multilayer-perceptron-numpy-version-

This is a 3 layer MLP Neural Network using only Numpy that can help us understand the principle of the neural network.

ai-tutorial-multiple-linear-regression

A basic multiple linear regression example. the training data was from mubaris

ai-tutorial-perceptron

This is a hard coding version of Sigmoid Perceptron. It is just for "Hello World" for the A.I beginners.

ai-tutorial-perceptrongui

This is a Perceptron with JavaFX GUI and error chart to observe the SGD training.

ai-tutorial-python-data-science-basic

This is a series of Python tutorial for basic data science.

ai-tutorial-tess4j-example

A Hello World of Tess4j Example

ai-tutorial-vanillarnn

This is a break down / hard coding version of an RNN with three timesteps and solving the binary addition problem.

bobatea

a toolset wrapper for Sklearn , pandas ... etc. for data analysis tasks and ML tasks in terminal.

chatgptcodecopilot

using chatGPT to help coding

convolution1d_experiment

A visual Experiment to explain signal filtering and convolution

cycle

Cycle.sh is a small bash script to auto-run the custom commands with looping in folders in linux-like system

handwrite_dataset_generator

This is a tool which helps to build handwrite digit dataset like MNIST. It can extract image patch(figure.2 ) from many printing forms(figure. 1), and build a 10 class dataset with labels.