Breast Cancer Prediction

Objective of the Project:

The aim of this machine learning project is to simulate a medical diagnosis scenario for educational purposes. We will assist healthcare professionals in making informed decisions by predicting whether a tumor is malignant or benign based on patient data. Although the data used in this project is real-world data sourced from Kaggle, it is intended solely for learning, and the insights gained will guide us in providing accurate and reliable predictions using a machine learning model.

I got the data from Kaggle

Data Description:

For this project, we utilized a dataset sourced from Kaggle, which contains various metrics about patients diagnosed with tumors. The dataset includes:

Radius Mean: The average radius of the tumor cells.

Texture Mean: The standard deviation of gray-scale values of the tumor.

Perimeter Mean: The mean size of the tumor perimeter.

Area Mean: The mean size of the tumor area.

Smoothness Mean: The mean smoothness of the tumor surface.

Compactness Mean: The mean compactness of the tumor.

Concavity Mean: The mean concavity of the tumor.

Concave Points Mean: The mean number of concave points on the tumor surface.

Symmetry Mean: The mean symmetry of the tumor.

Fractal Dimension Mean: The mean fractal dimension of the tumor.

Please note that while the dataset represents actual patient data, it has been anonymized for the purpose of this educational project.

Evaluation

In evaluating the performance of our machine learning model, which was built to predict whether a tumor is malignant or benign, we obtained the following metrics:

Accuracy: 0.98

Precision: 0.99 (for benign) / 0.97 (for malignant)

Recall: 0.98

The accuracy of the model indicates that 98% of the predictions made by the model are correct. The precision score reflects the proportion of true positive predictions among all positive predictions, with 99% precision for benign tumors and 97% for malignant tumors. The recall score, on the other hand, measures the model's ability to correctly identify true positive cases, with a recall of 98%. These metrics collectively demonstrate the high accuracy and reliability of the model in predicting tumor malignancy, making it a valuable tool for aiding healthcare professionals in decision-making.

License

Distributed under the MIT License. Click LICENSE.md for more information.

(back to top)

Contact

Akhin Abraham - twitter.com/akhinabr - theakhinabraham@gmail.com

Repository Link: https://github.com/theakhinabraham/cancer-prediction

(back to top)