Our project focused on using a dataset called the Breast Cancer Wisconsin (Diagnostic) dataset to develop a classification model for breast cancer. The main goal was to accurately predict whether breast tissue is malignant or benign, which is crucial for effective treatment.
We started by collecting the dataset, which consisted of features extracted from digitized images of breast masses. These features included various measurements related to the size, texture, and shape of the breast tissue. We then preprocessed the data by handling missing values, encoding categorical variables, and scaling the features.
Next, we selected the k-Nearest Neighbors (KNN) algorithm as our model. KNN works by comparing the features of a new sample with existing samples in the dataset to determine its class. We evaluated different distance metrics and feature scaling techniques to find the optimal settings for our model.
During the evaluation, we assessed the model's performance using metrics like accuracy. The results showed that our model achieved high accuracy in distinguishing between malignant and benign breast tissue. We also presented our findings using visualizations, such as charts and graphs, to enhance understanding.
While our project had limitations, such as the dataset size and class imbalance, we identified areas for improvement and future research. This includes exploring additional features and validating the model using external datasets.
In conclusion, our project highlighted the significance of early detection and accurate diagnosis in breast cancer. By leveraging the power of machine learning and developing a robust classification model, we aim to improve breast cancer diagnosis and ultimately enhance treatment outcomes.