1,720,961 research outputs found
Learning in Variable-Dimensional Spaces
No full textThis paper proposes a unified approach to learning in environments in which patterns can be represented in variable-dimension domains, which nicely includes the case in which there are missing features. The proposal is based on the representation of the environment by pointwise constraints that are shown to model naturally pattern relationships that come out in problems of information retrieval, computer vision, and related fields. The given interpretation of learning leads to capturing the truly different aspects of similarity coming from the content at different dimensions and the pattern links. It turns out that functions that process real-valued features and functions that operate on symbolic entities are learned within a unified framework of regularization that can also be expressed using the kernel machines mathematical and algorithmic apparatus. Interestingly, in the extreme cases in which only the content or only the links are available, our theory returns classic kernel machines or graph regularization, respectively. We show experimental results that provide clear evidence of the remarkable improvements that are obtained when both types of similarities are exploited on artificial and real-world benchmarks
Data Augmentation Techniques and Transfer Learning Approaches Applied to Facial Expressions Recognition Systems
Full text linkThe face expression is the first thing we pay attention to when we want to understand a person’s state of mind. Thus, the ability to recognize facial expressions in an automatic way is a very interesting research field. In this paper, because the small size of available training datasets, we propose a novel data augmentation technique that improves the performances in the recognition task. We apply geometrical transformations and build from scratch GAN models able to generate new synthetic images for each emotion type. Thus, on the augmented datasets we fine tune pretrained convolutional neural networks with different architectures. To measure the generalization ability of the models, we apply extra-database protocol approach, namely we train models on the augmented versions of training dataset and test them on two different databases. The combination of these techniques allows to reach average accuracy values of the order of 85\% for the InceptionResNetV2 model
Semantic-based regularization for learning and inference
No full textThis paper proposes a unified approach to learning from constraints, which integrates the ability of classical machine learning techniques to learn from continuous feature-based representations with the ability of reasoning using higher-level semantic knowledge typical of Statistical Relational Learning. Learning tasks are modeled in the general framework of multi-objective optimization, where a set of constraints must be satisfied in addition to the traditional smoothness regularization term. The constraints translate First Order Logic formulas, which can express learning-from-example supervisions and general prior knowledge about the environment by using fuzzy logic. By enforcing the constraints also on the test set, this paper presents a natural extension of the framework to perform collective classification. Interestingly, the theory holds for both the case of data represented by feature vectors and the case of data simply expressed by pattern identifiers, thus extending classic kernel machines and graph regularization, respectively. This paper also proposes a probabilistic interpretation of the proposed learning scheme, and highlights intriguing connections with probabilistic approaches like Markov Logic Networks. Experimental results on classic benchmarks provide clear evidence of the remarkable improvements that are obtained with respect to related approaches
Integrating Logic Knowledge into Graph Regularization: an application to image tagging
No full textThis paper studies how prior knowledge in form of First
Order Logic (FOL) clauses can be converted into a set of
continuous constraints. These constraints can be directly integrated
into a learning framework allowing to jointly learn
from examples and semantic knowledge. In particular, in
this paper we show how the constraints can be integrated
into a regularization schema working over discrete domains.
We consider tasks in which items are connected to each other
by given relationships, thus yielding a graph, whose nodes
correspond to the available objects. It is required to estimate
a set of functions defined on each node of the graph,
given a small set of labeled nodes for each function. The
FOL constraints enforce dependencies, resulting from the
FOL knowledge, among the values that the functions assume
over the nodes. The experimental results evaluate the
proposed technique on an image tagging task, showing how
the proposed approach provides a significantly higher tagging
accuracy than simple graph regularization. The experimental
results show how the selection of a proper conversion
process of the FOL clauses is fundamental in order to achieve
good results
Improved multi-level protein–protein interaction prediction with semantic-based regularization
Full text linkBackground: Protein-protein interactions can be seen as a hierarchical process occurring at three related levels: proteins bind by means of specific domains, which in turn form interfaces through patches of residues. Detailed knowledge about which domains and residues are involved in a given interaction has extensive applications to biology, including better understanding of the binding process and more efficient drug/enzyme design. Alas, most current interaction prediction methods do not identify which parts of a protein actually instantiate an interaction. Furthermore, they also fail to leverage the hierarchical nature of the problem, ignoring otherwise useful information available at the lower levels; when they do, they do not generate predictions that are guaranteed to be consistent between levels.
Results: Inspired by earlier ideas of Yip et al. (BMC Bioinformatics 10: 241, 2009), in the present paper we view the problem as a multi-level learning task, with one task per level (proteins, domains and residues), and propose a machine learning method that collectively infers the binding state of all object pairs. Our method is based on Semantic Based Regularization (SBR), a flexible and theoretically sound machine learning framework that uses First Order Logic constraints to tie the learning tasks together. We introduce a set of biologically motivated rules that enforce consistent predictions between the hierarchy levels.
Conclusions: We study the empirical performance of our method using a standard validation procedure, and compare its performance against the only other existing multi-level prediction technique. We present results showing that our method substantially outperforms the competitor in several experimental settings, indicating that exploiting the hierarchical nature of the problem can lead to better predictions. In addition, our method is also guaranteed to produce interactions that are consistent with respect to the protein-domain-residue hierarchy
Learning to Tag from Logic Constraints in Hyperlinked Environments
No full textThis paper presents a novel framework to integrate prior knowledge, represented as a collection of First Order Logic (FOL) clauses, into regularization over discrete domains. In particular, we consider tasks in which a set of items are connected to each other by given relationships yielding a graph, whose nodes correspond to the available objects, and it is required to estimate a set of functions defined on each node of the graph, given a small set of labeled nodes for each function. The available prior knowledge imposes a set of constraints among the function values. In particular, we consider background knowledge expressed as FOL clauses, whose predicates correspond to the functions and whose variables range over the nodes of the graph. These clauses can be converted into a set of constraints that can be embedded into a graph regularization schema. The experimental results evaluate the proposed technique on an image tagging task, showing how the proposed approach provides a significantly higher tagging accuracy than simple graph regularization. © 2011 IEEE
Experimental guidelines for semantic-based regularization
No full textThis paper presents a novel approach for learning with constraints called Semantic-Based Regularization. This paper shows how prior knowledge in form of First Order Logic (FOL) clauses, converted into a set of continuous constraints and integrated into a learning framework, allows to jointly learn from examples and semantic knowledge. A series of experiments on artificial learning tasks and application of text categorization in relational context will be presented to emphasize the benefits given by the introduction of logic rules into the learning process. © Springer International Publishing Switzerland 2014
Graph and Manifold Co-Regularization
No full textClassical foundations of Statistical Learning Theory rely on the assumption that the input patterns are independently and identically distributed. However, in many applications, the inputs, represented as feature vectors, are also embedded into a network of pairwise relations. Transductive approaches like graph regularization rely on the network topology without considering the feature vectors. Semi-supervised approaches like Manifold Regularization learn a function taking the feature vectors as input, while being smooth over the network connections. In this latter case, the connectivity information is processed at training time, but is still neglected during generalization, as the final classification decision takes only the feature vector representations as input. This paper presents and evaluates a model merging the advantages of graph regularization and kernel machines for transductive classification problems
Data Augmentation and Transfer Learning Approaches Applied to Facial Expressions Recognition
Full text linkThe face expression is the first thing we pay attention to when we want to understand a person’s state of mind. Thus, the ability to recognize facial expressions in an automatic way is a very interesting research field. In this paper, because the small size of available training datasets, we propose a novel data augmentation technique that improves the performances in the recognition task. We apply geometrical transformations and build from scratch GAN models able to generate new synthetic images for each emotion type. Thus, on the augmented datasets we fine tune pretrained convolutional neural networks with different architectures. To measure the generalization ability of the models, we apply extra-database protocol approach, namely we train models on the augmented versions of training dataset and test them on two different databases. The combination of these techniques allows to reach average accuracy values of the order of 85\% for the InceptionResNetV2 model
- …
