Manifold alignment is very prevalent in machine learning for extracting common latent space from multiple datasets. These algorithms generally aim to achieve higher alignment accuracies by preserving the original structure while ensuring closeness between manifolds. This paper proposes a novel semi-supervised manifold alignment method that combines in each manifold both global and local linear reconstructions. We preserve local structure through multiple manifold embedding methods. Moreover, we view manifold embedding methods as special forms of Principal Component Analysis (PCA) and thus, present a new penalty weight PCA approach to preserve noise-free global structure. Finally, a close-form solution is presented in manifold alignment. This method can concurrently match the pairwise correspondence and preserve both global and local structures of each dataset to obtain a latent lowdimensional space. Extensive experiments on manifold alignment prove that the propose method achieves significantly better alignment results than the comparative methods.

Patients, hospitals, sensors, researchers, providers, phones, and healthcare organisations are producing enormous amounts of data in both the healthcare and drug detection sectors. The real challenge in these sectors is to find, investigate, manage, and collect information from patients in order to make their lives easier and healthier, not only in terms of formulating new therapies and understanding diseases, but also to predict the results at earlier stages and make effective decisions. The volumes of data available in the fields of pharmacology, toxicology, and pharmaceutics are constantly increasing. These increases are driven by advances in technology, which allow for the analysis of ever-larger data sets. Big Data (BD) has the potential to transform drug development and safety testing by providing new insights into the effects of drugs on human health. However, harnessing this potential involves several challenges, including the need for specialised skills and infrastructure. In this survey, we explore how BD approaches are currently being used in the pharmacology, toxicology, and pharmaceutics fields; in particular, we highlight how researchers have applied BD in pharmacology, toxicology, and pharmaceutics to address various challenges and establish solutions. A comparative analysis helps to trace the implementation of big data in the fields of pharmacology, toxicology, and pharmaceutics. Certain relevant limitations and directions for future research are emphasised. The pharmacology, toxicology, and pharmaceutics fields are still at an early stage of BD adoption, and there are many research challenges to be overcome, in order to effectively employ BD to address specific issues.