Research Article
[Retracted] Machine Learning-Based Gynecologic Tumor Diagnosis and Its Postoperative Incisional Infection Influence Factor Analysis
| Input: dataset | | Output: attention matrix O, prediction result Prediction | | (1) | Random initialization matrix | | (2) | Embedding encoding of data datasets | | (3) | for i in n do | | (4) | | | (5) | | | (6) | | | (7) | End for | | (8) | Do attention to each with | | (9) | For i in n do | | (10) | For j in n do | | (11) | , d is the dim of q and k | | (12) | End for | | (13) | End for | | (14) | Normalize the after attention | | (15) | | | (16) | For i in n do | | (17) | For j in n do | | (18) | | | (19) | End for | | (20) | End for | | (21) | Initialize the adaptive neural network weights W and bias b | | (22) | The matrix obtained after attention is input to the neural network for training | | (23) | Get the final prediction result Prediction = Softmax(∙ O + b) |
|