/Spam-Mail-Detection-ML

Primary LanguagePython

Model Execution Guide

Installations

You'll need to set up a Python environment and install libraries commonly used in machine learning.

You don't need all of them to run this file. The following command is all you need:

pip install numpy pandas scikit-learn

Algo! :

Step-by-Step Algorithm for Running the ML Model

1. Import Libraries

  • Import necessary libraries for data manipulation, machine learning, and evaluation.

2. Load the Dataset

  • Load the dataset from a CSV file into a pandas DataFrame.

3. Handle Missing Values

  • Replace null values in the DataFrame with empty strings.

4. Encode Labels

  • Encode the 'Category' column labels: 'spam' as 0 and 'ham' as 1.

5. Separate Features and Labels

  • Separate the feature (messages) and the target (labels).

6. Split the Data

  • Split the data into training and test sets.

7. Transform Text Data to Feature Vectors

  • Transform the text data into TF-IDF feature vectors.

8. Train the Logistic Regression Model

  • Train a logistic regression model using the training data.

9. Make Predictions

  • Use the trained model to make predictions on the test data.

10. Evaluate the Model

  • Evaluate the model's performance using a confusion matrix and accuracy score.

Output :

1 .Confusion Matrix

   147  49
    0   1197
    
    True Negatives (TN)  : 147
    False Positives (FP) : 49
    False Negatives (FN) : 0
    True Positives (TP)  : 1197

2 .Accuracy Score

0.964824120603015

An accuracy score of 96.48% means that the model correctly predicted the category of emails about 96.48% of the time.