Nature language inference (NLI) task is a predictive task of determining the inference relationship of a pair of natural language sentences. With the increasing popularity of NLI, many state-of-the-art predictive models have been proposed with impressive performances. However, several works have noticed the statistical irregularities in the collected NLI data set that may result in an over-estimated performance of these models and proposed remedies. In this paper, we further investigate the statistical irregularities, what we refer as confounding factors, of the NLI data sets. With the belief that some NLI labels should preserve under swapping operations, we propose a simple yet effective way (swapping the two text fragments) of evaluating the NLI predictive models that naturally mitigate the observed problems. Further, we continue to train the predictive models with our swapping manner and propose to use the deviation of the model's evaluation performances under different percentages of training text fragments to be swapped to describe the robustness of a predictive model. Our evaluation metrics leads to some interesting understandings of recent published NLI methods. Finally, we also apply the swapping operation on NLI models to see the effectiveness of this straightforward method in mitigating the confounding factor problems in training generic sentence embeddings for other NLP transfer tasks.
Neural network has obvious advantages on dealing with the uncertain problems with a huge amount of data. Some equipment’s faults performance of such characteristics that the date is abundant, and failure phenomenon is not explicit and uncertain. Then, it leads to being hard to diagnose the faults through the traditional diagnosis method in a short time. This paper will analysis the data feature and then build a model to deal with the qualitative data attributes in order that the BP network can use it smoothly. Calculation result shows that using this method, fault diagnosis can be simply and quickly. The paper also provides a new kind of composite way to figure out fault positions for the front-line operators based on experts’experience knowledge but not on measurement signals.
Grey system theory has obvious advantage on handling “poor information and small sample data” problems. Some TV seeker’s faults performance of such characteristics that the date is less and not integrity, and failure phenomenon is not explicit. So it comes to be difficult to diagnose the faults at the Basic level by a quick way. This paper means to build a model using the general gray relational analysis in order to establish contact between the fault parameter names and failure phenomenon. Then by using gray relational analysis on the fault records in the database and checking the model with another data. Calculation result shows that the method can judge fault position simply and quickly. The paper provides a kind of relatively reliable quantitative analysis basis for the fault diagnosis and locating faults.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.