Pattern-Recognition-Assignment1

How to run this project

• Step1: Download this project or just clone it via git clone. The repository is here

• Step2: Here are some runtime environment you should already have installed before running this project. All of these should be prepared if you don't have them all installed correctly.

  • node.js ( npm should be installed along with node.js )
  • typescript ( We truely recommend installing typescript via npm: npm i typescript -g )
  • ts-node may have access to simplify the procedure running this project. However, run compiled js files on node.js is always a good solution.
  • Angular cli npm i -g @angular/cli is also required if you want to run the data visualization project inside the folder 'LogisticRegressionVisualization'. We also provide a compiled version of this project which you can run it in browsers like Chrome. But a local server is required instead( We recommend the live server extension for vscode ).

• step3: Compile this project

  • Before compile this project, you must run npm i to install all the dependencies which this project use.
  • Use the command tsc to compile all .ts files into .js files. All the js file will be waiting for you inside the 'dist' folder.
  • Alternatively, you could leave this job to ts-node and skip this step.

• step4: Run this project

  To run this project use the command node dist/index.js or use ts-node src/index.ts. Use the method whichever you like to run this project.

Structure of this project

• src

  All the source files are in this folder.

  • index.ts is the entrance of the program. We load iris data and train our logistic regression model here. Besides, the model will be saved at myLogisticRegression1.json. You can load it to do prediction next time.
  • shuffle.ts shuffles the iris data and saves it in iris_processed.json.
  • Just ignore test.ts.
  • The validate folder includes a validation program to test whether our program is running correctly.
  • The maths folder contains the mathematic tools we used to implement the logistic regression model including optimizers and probablity functions.
  • The logistic folder is the very place where our logistic regression algrithm lies.

• dist

  • This folder contains all of the compiled .js files.

• Pattern-Recognition-Assignment1-Visualization

  • This folder is a project which is a undergoing logistic regression visualization application.
  • To run this project use npm i and the input ng serve --open in the command line.
  • For more detailed information, reference the README.md file inside this folder.

• Others

  • Just leave them alone.