Learning with similarity functions : a tensor-based framework
Machine learning algorithms are typically designed to deal with data represented as vectors. Several major applications, however, involve multi-way data, such as video sequences and multi-sensory arrays. In those cases, tensors endow a more consistent way to capture multi-modal relations, which may...
| Main Authors: | , , , |
|---|---|
| Other Authors: | |
| Format: | Journal Article |
| Language: | English |
| Published: |
2021
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/150714 |
| _version_ | 1826117810545229824 |
|---|---|
| author | Ragusa, Edoardo Gastaldo, Paolo Zunino, Rodolfo Cambria, Erik |
| author2 | School of Computer Science and Engineering |
| author_facet | School of Computer Science and Engineering Ragusa, Edoardo Gastaldo, Paolo Zunino, Rodolfo Cambria, Erik |
| author_sort | Ragusa, Edoardo |
| collection | NTU |
| description | Machine learning algorithms are typically designed to deal with data represented as vectors. Several major applications, however, involve multi-way data, such as video sequences and multi-sensory arrays. In those cases, tensors endow a more consistent way to capture multi-modal relations, which may be lost by a conventional remapping of original data into a vector representation. This paper presents a tensor-oriented machine learning framework, and shows that the theory of learning with similarity functions provides an effective paradigm to support this framework. The proposed approach adopts a specific similarity function, which defines a measure of similarity between a pair of tensors. The performance of the tensor-based framework is evaluated on a set of complex, real-world, pattern-recognition problems. Experimental results confirm the effectiveness of the framework, which compares favorably with state-of-the-art machine learning methodologies that can accept tensors as inputs. Indeed, a formal analysis proves that the framework is more efficient than state-of-the-art methodologies also in terms of computational cost. The paper thus provides two main outcomes: (1) a theoretical framework that enables the use of tensor-oriented similarity notions and (2) a cognitively inspired notion of similarity that leads to computationally efficient predictors. |
| first_indexed | 2024-10-01T04:33:26Z |
| format | Journal Article |
| id | ntu-10356/150714 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T04:33:26Z |
| publishDate | 2021 |
| record_format | dspace |
| spelling | ntu-10356/1507142021-06-08T03:10:00Z Learning with similarity functions : a tensor-based framework Ragusa, Edoardo Gastaldo, Paolo Zunino, Rodolfo Cambria, Erik School of Computer Science and Engineering Engineering::Computer science and engineering Tensor Data Similarity Functions Machine learning algorithms are typically designed to deal with data represented as vectors. Several major applications, however, involve multi-way data, such as video sequences and multi-sensory arrays. In those cases, tensors endow a more consistent way to capture multi-modal relations, which may be lost by a conventional remapping of original data into a vector representation. This paper presents a tensor-oriented machine learning framework, and shows that the theory of learning with similarity functions provides an effective paradigm to support this framework. The proposed approach adopts a specific similarity function, which defines a measure of similarity between a pair of tensors. The performance of the tensor-based framework is evaluated on a set of complex, real-world, pattern-recognition problems. Experimental results confirm the effectiveness of the framework, which compares favorably with state-of-the-art machine learning methodologies that can accept tensors as inputs. Indeed, a formal analysis proves that the framework is more efficient than state-of-the-art methodologies also in terms of computational cost. The paper thus provides two main outcomes: (1) a theoretical framework that enables the use of tensor-oriented similarity notions and (2) a cognitively inspired notion of similarity that leads to computationally efficient predictors. 2021-06-08T03:09:59Z 2021-06-08T03:09:59Z 2019 Journal Article Ragusa, E., Gastaldo, P., Zunino, R. & Cambria, E. (2019). Learning with similarity functions : a tensor-based framework. Cognitive Computation, 11(1), 31-49. https://dx.doi.org/10.1007/s12559-018-9590-9 1866-9956 0000-0002-3030-1280 https://hdl.handle.net/10356/150714 10.1007/s12559-018-9590-9 2-s2.0-85053384656 1 11 31 49 en Cognitive Computation © 2018 Springer Science Business Media, LLC, part of Springer Nature. All rights reserved. |
| spellingShingle | Engineering::Computer science and engineering Tensor Data Similarity Functions Ragusa, Edoardo Gastaldo, Paolo Zunino, Rodolfo Cambria, Erik Learning with similarity functions : a tensor-based framework |
| title | Learning with similarity functions : a tensor-based framework |
| title_full | Learning with similarity functions : a tensor-based framework |
| title_fullStr | Learning with similarity functions : a tensor-based framework |
| title_full_unstemmed | Learning with similarity functions : a tensor-based framework |
| title_short | Learning with similarity functions : a tensor-based framework |
| title_sort | learning with similarity functions a tensor based framework |
| topic | Engineering::Computer science and engineering Tensor Data Similarity Functions |
| url | https://hdl.handle.net/10356/150714 |
| work_keys_str_mv | AT ragusaedoardo learningwithsimilarityfunctionsatensorbasedframework AT gastaldopaolo learningwithsimilarityfunctionsatensorbasedframework AT zuninorodolfo learningwithsimilarityfunctionsatensorbasedframework AT cambriaerik learningwithsimilarityfunctionsatensorbasedframework |