Your dataset seems a bit problematic since it is small and niche, though it is probably too late for that. "Most of the women in the dataset are under the age of 40 and the average age is a little less than 27" worries me because this may not be representative of the population as a whole and cancer usually appears in older people. However, your proposed overfitting technique seems pretty valid since you are accounting for false positives and false negatives that might come about because of your small dataset. Maybe you could find more data and combine it with your current data! Also, I would have liked more specifics on the features such as what data type and how they contribute to your predictions. Great job on everything though, very impressive!
Your dataset seems a bit problematic since it is small and niche, though it is probably too late for that. "Most of the women in the dataset are under the age of 40 and the average age is a little less than 27" worries me because this may not be representative of the population as a whole and cancer usually appears in older people. However, your proposed overfitting technique seems pretty valid since you are accounting for false positives and false negatives that might come about because of your small dataset. Maybe you could find more data and combine it with your current data! Also, I would have liked more specifics on the features such as what data type and how they contribute to your predictions. Great job on everything though, very impressive!