Ball_Classification

A beginner ML project

This is one of the simplest machine learning project to get started.

What is the project about?

This project predicts the lable of given input data Whether it is a Tennis Ball or Cricket Ball. This project also demonstrate the use of pickle file. To know more about pickle file in detail click here

How the dataset looks like?

Weight	Surface	Label
35	Rough	Tennis
47	Rough	Tennis
90	Smooth	Cricket
48	Rough	Tennis

Understanding the project:

• First of all we need to read the dataset from the given Ball_Dataset.csv file

  • To read csv file use the read_csv() function from pandas.

• To train a ML model we need numerical data, but here we have string data.

  • So, convert the string data into numerical data using the function factorize().
  • For more details refer the links stackoverflow-factorize() or pydata-factorize().
  • After conversion the dataset looks like as shown below:

Weight	Surface	Label
35	0	0
47	0	0
90	1	1
48	0	1

• After converting string data into numeric form, you need to seperate it for Training and Testing

  • For this use the function train_test_split() from sklearn.model_selection. This function splits the data into two parts as per the given ratio.
  • For more understanding of this function click here.

• Now to train the data using the training dataset:

  • Create an object of class DecisionTreeClassifier() and load the training data to train the model using the fit() function.
  • To learn more about DecisionTreeClassifier() and fit() click here.

• After training its time to test the model:

  • use the predict() function of the DecisionTreeClassifier() class.
  • This will give the lable for given testing data.
  • To know more about predict() function click here.

How to execute program?

• Here the program is divided into 2 modules train ans test
• To execute Inbuilt_argparse.py to train model, use the below command:

> python InBuilt_argparse.pr -tr
# or youcan use
> python InBuilt_argparse.py --train

• To execute Inbuilt_argparse.py to test model, use the below command:

> python InBuilt_argparse.pr -te
# or you can use
> python InBuilt_argparse.py --test

• To execute Inbuilt_streamlit.py, use the below command in terminal:

> streamlit run InBuilt_streamlit.py

If you have any doub or problem then feel free to raise an issue

If you find this repository helpful ⭐ this repository and share with your friends.