/ML-letter-recognition

A couple of machine learning algorithms utilising the UCI dataset for letter recognition.

Primary LanguagePython

Machine Learning Methods for letter recognition

A couple of machine learning algorithms utilising the UCI dataset for letter recognition

Getting Started

Install dependencies

Prerequisites (for everyone)

The project requires the following major dependencies:

  • python 2.7
  • pip

To install dependencies:

  1. Check your python version.
python -V
  1. Install all the dependencies
pip install -r requirements.txt

Run Python scripts

python <file-name>

Concept

We are apply the different machine learning algorithm concepts on the UCI Letter Recognition Data Set .

Given the 16 attributes, are we able to accurately predict the letter category as one of the 26 capital letters in the English alphabet.

The whole concept of this project is that we are looking to identify:

  1. Which attributes are more valuable to solving our problem?

  2. Which attributes are least valuable to solving our problem?

  3. What are the list of attributes that create the highest accuracy count?

  4. What are the list of attributes that create the lowest accuracy count?

  5. What model is most suitable for this problem?

  6. What model is least suitable for this problem?