More Like this

1. Issues of Lexicon in South-Central Tibeto-Burman (Kuki-Chin)

   https://doi.org/10.5070/H922153742  

   Van_Bik, Kenneth ( July 2023 , Himalayan Linguistics)

   This paper examines how the lexicon is organized in a typical South Central language. Items like nouns, verbs, and adverbial expressions belong to open classes; pronominals, demonstratives, numerals, quantifiers, interjections, and onomatopoetic words form closed classes. Middle markers, case markers, directionals, tense/aspect markers, valence-changing elements, verbal classifiers, elaborate expressions, and reduplicative patterns are treated as bound elements. 

   more » « less
2. Variation in pronominal indexing: Lexical stipulation vs. referential properties in Alor-Pantar languages

   https://doi.org/10.1075/sl.38.1.02fed  

   Fedden, Sebastian ; Brown, Dunstan ; Kratochvíl, František ; Robinson, Laura C. ; Schapper, Antoinette ( January 2014 , Studies in Language)

   We examine the role of referential properties and lexical stipulation in three closely related languages of eastern Indonesia, the Alor-Pantar languages Abui, Kamang, and Teiwa. Our focus is on the continuum along which event properties (e.g. volitionality, affectedness) are highly important at one extreme or play virtually no role at the other. These languages occupy different points along this continuum. In Abui, event semantics play the greatest role, while in Teiwa they play the smallest role (the lexical property animacy being dominant in the formation of verb classes). Kamang occupies an intermediate position. Teiwa has conventionalised the relation between a verb and its class along the lines of animacy so that classes become associated with the animacy value of the objects with which the verbs in a given class typically occur. Paying attention to a lexical property like animacy, in contrast with event properties, has meant greater potential for arbitrary classes to emerge. 

   more » « less
3. Open Set Text Classification using Convolutional Neural Networks

   Prakhya, Sridhama ; Venkataram, Vinodini ; Kalita, Jugal ( December 2017 , International Conference on Natural Language Processing, 2017)

   In a closed world setting, classifiers are trained on examples from a number of classes and tested with unseen examples belonging to the same set of classes. However, in most real-world scenarios, a trained classifier is likely to come across novel examples that do not belong to any of the known classes. Such examples should ideally be categorized as belonging to an unknown class. The goal of an open set classifier is to anticipate and be ready to handle test examples of classes unseen during training. The classifier should be able to declare that a test example belongs to a class it does not know, and possi- bly, incorporate it into its knowledge as an example of a new class it has encoun- tered. There is some published research in open world image classification, but open set text classification remains mostly un- explored. In this paper, we investigate the suitability of Convolutional Neural Net- works (CNNs) for open set text classifi- cation. We find that CNNs are good fea- ture extractors and hence perform better than existing state-of-the-art open set clas- sifiers in smaller domains, although their open set classification abilities in general still need to be investigated. 

   more » « less
4. Theoretical Insights Into Multiclass Classification: A high-dimensional Asymptotic View

   Thrampoulidis, Christos ( January 2020 , Neural Information Processing Systems (NeuRIPS 2020))

   null (Ed.)

   Contemporary machine learning applications often involve classification tasks with many classes. Despite their extensive use, a precise understanding of the statistical properties and behavior of classification algorithms is still missing, especially in modern regimes where the number of classes is rather large. In this paper, we take a step in this direction by providing the first asymptotically precise analysis of linear multiclass classification. Our theoretical analysis allows us to precisely character- ize how the test error varies over different training algorithms, data distributions, problem dimensions as well as number of classes, inter/intra class correlations and class priors. Specifically, our analysis reveals that the classification accuracy is highly distribution-dependent with different algorithms achieving optimal per- formance for different data distributions and/or training/features sizes. Unlike linear regression/binary classification, the test error in multiclass classification relies on intricate functions of the trained model (e.g., correlation between some of the trained weights) whose asymptotic behavior is difficult to characterize. This challenge is already present in simple classifiers, such as those minimizing a square loss. Our novel theoretical techniques allow us to overcome some of these chal- lenges. The insights gained may pave the way for a precise understanding of other classification algorithms beyond those studied in this paper. 

   more » « less
5. Tag N’ Train: a technique to train improved classifiers on unlabeled data

   https://doi.org/10.1007/JHEP01(2021)153  

   Amram, Oz ; Suarez, Cristina Mantilla ( January 2021 , Journal of High Energy Physics)

   A bstract There has been substantial progress in applying machine learning techniques to classification problems in collider and jet physics. But as these techniques grow in sophistication, they are becoming more sensitive to subtle features of jets that may not be well modeled in simulation. Therefore, relying on simulations for training will lead to sub-optimal performance in data, but the lack of true class labels makes it difficult to train on real data. To address this challenge we introduce a new approach, called Tag N’ Train (TNT), that can be applied to unlabeled data that has two distinct sub-objects. The technique uses a weak classifier for one of the objects to tag signal-rich and background-rich samples. These samples are then used to train a stronger classifier for the other object. We demonstrate the power of this method by applying it to a dijet resonance search. By starting with autoencoders trained directly on data as the weak classifiers, we use TNT to train substantially improved classifiers. We show that Tag N’ Train can be a powerful tool in model-agnostic searches and discuss other potential applications. 

   more » « less