Edge-Based Convolutional Neural Network for Improving Breast Cancer Prediction Performance

Table 1

Comparative analysis of previous research studies to the proposed work.

Researches	Technology used	Key benefit	Performance factor	Storage space factor	Accuracy factor	Issues and challenges
[4]	CNN	Efficient image classification	No	No	No	Time consumption and space consumption need to be reduced
[7]	CNN and SVM	Fast performance	Yes	No	No	The issue of space consumption and accuracy not resolved
[12]	Random forest algorithm	Handling missing values, no feature scaling required, and less impacted by noise	No	No	No	Complexity and long training period
[22]	Genetic algorithm	It provides good quality solutions in a less time	Yes	No	No	Does not provide the optimal solution
[26]	Bayesian logistic regression	It provides better result that is unbiased, with lower variances	No	No	Yes	Logistic regression is capable of predicting a categorical outcome
[27]	Ensemble convolution neural networks	It offers increased flexibility	Yes	No	No	Does not consider the space and accuracy
[29]	Unsupervised feature extraction algorithm	Ideal to explore raw and unknown data	Yes	No	No	There is a lack of accuracy due to unavailability of labels
Proposed work	CNN and edge detection		Yes	Yes	Yes	The integration of multiple technologies is quite challenging