Toward a Self-Supervised Architecture for Semen Quality Prediction Using Environmental and Lifestyle Factors

Abstract

Male fertility has been seen to be declining, prompting for more effective and accessible means of its assessment. AI methods have been effective towards predicting semen quality through a questionnaire-based information source comprising a selection of factors from the medical literature which have been seen to influence semen quality. Prior work has seen the application of supervised learning towards the prediction of semen quality, but since supervised learning hinges on the provision of data class labels it can be said to depend on an external intelligence intervention, which can translate towards further costs and resources in practical settings. In contrast, unsupervised learning methods partition data into clusters and groups based on an objective function and do not rely on the provision of class labels, and can allow for a fully automated flow of a prediction platform. In this paper we apply three unsupervised learning models with different model architectures, namely: Gaussian mixture model (GMM), K-means and spectral clustering (SC), alongside low dimensional embedding methods which include sparse autoencoder (SAE), principal component analysis (PCA) and robust principal component analysis (RPCA). The best results were obtained with a combination of the sparse autoencoder and the SC algorithm, which was likely due to its nonspecific and arbitrary cluster shape assumption. Further work would now involve the exploration of similar unsupervised learning algorithms with a similar framework to the SC to investigate the extent to which various clusters can be learned with maximal accuracy.

Traffic Marking Extraction Algorithm Based on Image and Point Cloud Data

“…Next, on the ground of the difference size at both ends of the edge, a threshold is artificially set, which is mainly used to distinguish the difference in grayscale values between edges and non edges. When the difference between the two ends of the edge is greater than this threshold, it indicates that the position is an edge pixel and can be extracted [20] ( , ) ( , ) ( , ) In edge detection on the ground of the Canny operator, the first step is to perform Gaussian filtering, which mainly uses normally distributed weights to weight and average the image, to smooth the image, decrease image noise and details. Next, it performs gradient calculation of the image, where the gradient of the image represents the edge intensity and direction at each pixel point in the image.…”

Real-Time Management of Coal Mine Underground Shield Machine Digging Speed Based on Improved Residual Neural Networks

“…The neural network chosen for the study is the eXtreme gradient boosting (XGBoost) algorithm. XGBoost algorithm is an improved decision tree (DT) algorithm, which can deal with a variety of complex and irregular data [20][21][22]. The surrounding rock type (SRT) prediction model constructed by the study is shown in Figure 1.…”

The Application of Improved Scale Invariant Feature Transformation Algorithm in Facial Recognition

“…The application fields of FR models are very extensive, which can be applied in security fields such as access control systems, identity verification, and crime investigation. In addition, FR can also be applied to human-computer interaction and has a wide range of applications [20]. The traditional FR model is shown in Fig.…”