Background Drug‒drug interactions (DDIs) are reactions between two or more drugs, i.e., possible situations that occur when two or more drugs are used simultaneously. DDIs act as an important link in both drug development and clinical treatment. Since it is not possible to study the interactions of such a large number of drugs using experimental means, a computer-based deep learning solution is always worth investigating. We propose a deep learning-based model that uses twin convolutional neural networks to learn representations from multimodal drug data and to make predictions about the possible types of drug effects. Results In this paper, we propose a novel convolutional neural network algorithm using a Siamese network architecture called CNN-Siam. CNN-Siam uses a convolutional neural network (CNN) as a backbone network in the form of a twin network architecture to learn the feature representation of drug pairs from multimodal data of drugs (including chemical substructures, targets and enzymes). Moreover, this network is used to predict the types of drug interactions with the best optimization algorithms available (RAdam and LookAhead). The experimental data show that the CNN-Siam achieves an area under the precision-recall (AUPR) curve score of 0.96 on the benchmark dataset and a correct rate of 92%. These results are significant improvements compared to the state-of-the-art method (from 86 to 92%) and demonstrate the robustness of the CNN-Siam and the superiority of the new optimization algorithm through ablation experiments. Conclusion The experimental results show that our multimodal siamese convolutional neural network can accurately predict DDIs, and the Siamese network architecture is able to learn the feature representation of drug pairs better than individual networks. CNN-Siam outperforms other state-of-the-art algorithms with the combination of data enhancement and better optimizers. But at the same time, CNN-Siam has some drawbacks, longer training time, generalization needs to be improved, and poorer classification results on some classes.

Deep-learning based industrial defect detection methods play an increasingly important role in intelligent manufacturing, as they provide compelling benefits in reducing the cost spent on manual product inspection and meanwhile, improving inspection accuracy and efficiency. They have been widely used in various manufacturing and O&M applications such as automated inspection, smart patrol and quality controlling. This survey aims to make a comprehensive introduction of industrial defect detection, which mainly spans its definition, difficulties, challenges, mainstream methods, open datasets and evaluation protocols, so as to help researchers gather a quick and broad understanding. Specifically, we firstly introduce some background knowledge. Secondly, based on the difference of the provided annotations of different datasets in practical scenarios, we categorize most methods into three task settings: known defects, unknown defects, and few-shot defects. We give more details over these methods and illustrate a thorough analysis. We expound the connections between different algorithms and actual demands to provide a clear image of how different algorithms evolve. Thirdly, this paper summarizes some useful strategies that can effectively improve defect detection performance. Finally, based on our understanding of this area, we conclude several limitations of existing methods in practical applications as well as several directions of future research that embrace further efforts.