Skip to content

RockDevelop/PokemonAI

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

51 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

PokemonAI - Machine Learning Final Project

File Description

pokemon.py - This is a Python script meant to be run with command line inputs to choose which model you wish to run. You have the choice to run a decision tree model or a neural network model. The model will predict based on that stats and body type of a Pokemon what type it will be. The ouput will be the accuracy of the training followed by the accuracy of the testing. There are also visual aids to see how the model is performing.

data.csv - This is a data set from Kaggle that contains information on the first VI generations of pokemon games. The data is in the order below

Number Name Type 1 Type 2 Stats Total HP Attack Defense Sp Attack Sp Defense Speed Generation isLegendary Color hasGender Percent Male Egg Group 1 Egg Group 2 hasMegaEvolution Height Meters Weight Kilograms Catch Rate Body Type

Required Libraries

  • Tensorflow - Backbone of the machine learning models
pip install tensorflow
  • Keras - Used to create, train, and test the neural network model
pip install keras
  • Matplotlib - Used to plot and show the graphs related to the models
pip install -U matplotlib
  • Scikit-learn - Used to create, train, and test the decision tree model
pip install -U scikit-learn
  • Numpy - Used for data manipulation and ease of access to more complex data structres
pip install numpy
  • Wordcloud - Used for the visualization of the neural network predictions
pip install wordcloud

How to run

You have two choices when running the model. You are able to choose if you want to run the neural network model, or the decision tree model. The command line arguments are below.

  • Neural Network
python pokemon.py -model nn
  • Decision Tree
python pokemon.py -model dt

About

Final project for AI 2

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 2

  •  
  •  

Languages