PT116910A - TWO-DIMENSIONAL MULTI-CONVOLUTIONAL CARE UNIT FOR MULTIVARIABLE TIME SERIES ANALYSIS - Google Patents
TWO-DIMENSIONAL MULTI-CONVOLUTIONAL CARE UNIT FOR MULTIVARIABLE TIME SERIES ANALYSISInfo
- Publication number
- PT116910A PT116910A PT116910A PT11691020A PT116910A PT 116910 A PT116910 A PT 116910A PT 116910 A PT116910 A PT 116910A PT 11691020 A PT11691020 A PT 11691020A PT 116910 A PT116910 A PT 116910A
- Authority
- PT
- Portugal
- Prior art keywords
- convolutional
- dimensional
- block
- attention
- time series
- Prior art date
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Molecular Biology (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
- Image Generation (AREA)
Abstract
É, PORTANTO, UM OBJECTIVO DA PRESENTE INVENÇÃO UMA UNIDADE DE ATENÇÃO MULTI-CONVOLUCIONAL BI-DIMENSIONAL - 2D - A SER APLICADA NO DESEMPENHO DE ANÁLISE DE DADOS TRI-DIMENSIONAL - 3D - MTS, DE DADOS DE ENTRADA (1) COM PROPRIEDADES CÍCLICAS, UTILIZANDO UMA ARQUITECTURA RNN. ESTA UNIDADE É CAPAZ DE CONSTRUIR UM VECTOR DE ATENÇÃO INDEPENDENTE Α POR CADA VARIÁVEL DO MTS UTILIZANDO OPERAÇÕES CONVOLUCIONAIS 2D PARA CAPTURAR A IMPORTÂNCIA DE UMA ETAPA TEMPORAL DENTRO DOS SEGMENTOS CIRCUNDANTES E DA ÁREA DE ETAPAS TEMPORAIS. PARA ESSE PROPÓSITO, A UNIDADE DE ATENÇÃO BIDIMENSIONAL É COMPOSTA POR UM BLOCO DE DIVISÃO (2), UM BLOCO DE ATENÇÃO (3), UM BLOCO DE CONCATENAÇÃO (4) E UM BLOCO DE ESCALONAMENTO (5).IT IS, THEREFORE, AN OBJECTIVE OF THE PRESENT INVENTION A TWO-DIMENSIONAL - 2D - MULTI-CONVOLUTIONAL ATTENTION UNIT TO BE APPLIED IN THE PERFORMANCE OF THREE-DIMENSIONAL - 3D - MTS DATA ANALYSIS, OF INPUT DATA (1) WITH CYCLIC PROPERTIES, USING AN RNN ARCHITECTURE. THIS UNIT IS CAPABLE OF CONSTRUCTING AN INDEPENDENT ATTENTION VECTOR Α FOR EACH MTS VARIABLE USING CONVOLUTIONAL 2D OPERATIONS TO CAPTURE THE IMPORTANCE OF A TEMPORAL STEP WITHIN THE SURROUNDING SEGMENTS AND THE TEMPORAL STEPS AREA. FOR THIS PURPOSE, THE TWO-DIMENSIONAL ATTENTION UNIT IS COMPOSED OF A SPLITTING BLOCK (2), A ATTENTION BLOCK (3), A CONCATENATION BLOCK (4) AND A STEERING BLOCK (5).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PT116910A PT116910B (en) | 2020-11-30 | 2020-11-30 | TWO-DIMENSIONAL MULTI-CONVOLUTIONAL CARE UNIT FOR MULTIVARIABLE TIME SERIES ANALYSIS |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PT116910A PT116910B (en) | 2020-11-30 | 2020-11-30 | TWO-DIMENSIONAL MULTI-CONVOLUTIONAL CARE UNIT FOR MULTIVARIABLE TIME SERIES ANALYSIS |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| PT116910A true PT116910A (en) | 2022-05-30 |
| PT116910B PT116910B (en) | 2022-09-21 |
Family
ID=81827394
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PT116910A PT116910B (en) | 2020-11-30 | 2020-11-30 | TWO-DIMENSIONAL MULTI-CONVOLUTIONAL CARE UNIT FOR MULTIVARIABLE TIME SERIES ANALYSIS |
Country Status (1)
| Country | Link |
|---|---|
| PT (1) | PT116910B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017155660A1 (en) * | 2016-03-11 | 2017-09-14 | Qualcomm Incorporated | Action localization in sequential data with attention proposals from a recurrent network |
| US20170262995A1 (en) * | 2016-03-11 | 2017-09-14 | Qualcomm Incorporated | Video analysis with convolutional attention recurrent neural networks |
| US20200334334A1 (en) * | 2019-04-18 | 2020-10-22 | Salesforce.Com, Inc. | Systems and methods for unifying question answering and text classification via span extraction |
-
2020
- 2020-11-30 PT PT116910A patent/PT116910B/en active IP Right Grant
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017155660A1 (en) * | 2016-03-11 | 2017-09-14 | Qualcomm Incorporated | Action localization in sequential data with attention proposals from a recurrent network |
| US20170262995A1 (en) * | 2016-03-11 | 2017-09-14 | Qualcomm Incorporated | Video analysis with convolutional attention recurrent neural networks |
| US20200334334A1 (en) * | 2019-04-18 | 2020-10-22 | Salesforce.Com, Inc. | Systems and methods for unifying question answering and text classification via span extraction |
Also Published As
| Publication number | Publication date |
|---|---|
| PT116910B (en) | 2022-09-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Daniely | Depth separation for neural networks | |
| SG10201707700WA (en) | Performing Kernel Striding In Hardware | |
| RU2018121280A (en) | Cryptographic device adapted for calculating the target block cipher | |
| Neshatpour et al. | Accelerating big data analytics using fpgas | |
| PT116910A (en) | TWO-DIMENSIONAL MULTI-CONVOLUTIONAL CARE UNIT FOR MULTIVARIABLE TIME SERIES ANALYSIS | |
| Pantu | Online learning: the role of academic self-efficacy in creating academic flow | |
| Lee et al. | Imaging of the finger vein and blood flow for anti-spoofing authentication using a laser and a MEMS scanner | |
| Merca | On families of linear recurrence relations for the special values of the Riemann zeta function | |
| Durdiev et al. | The local solvability of a problem of determining the spatial part of a multidimensional kernel in the integro-differential equation of hyperbolic type | |
| Liu | A refinement of an equivalent form of a Gerretsen inequality | |
| KR950024083A (en) | Neural network and how to operate it | |
| Miller | Finding all monomials in a polynomial ideal | |
| Duarte et al. | A short proof of a famous combinatorial identity | |
| Kishka et al. | A solution of system of linear matrix equations by matrix of matrices | |
| Keefe | English Puritanism as represented by John Milton | |
| Gaul et al. | “It ain’t what you do, it’s the way that you do it”: does obesity affect perceptual motor control ability of adults on the speed and accuracy of a discrete aiming task? | |
| Rabea | The Role of Political Thought in the Japanese Renaissance: A Study in the Ideas of Shiga Shigitaka (1863-1927) | |
| Alice et al. | Druggability assessment algorithm based on Composition, Transition and Distribution descriptors and publicly available predictive tools | |
| Johnson et al. | Reply to Broxterman, Richardson, and Amann | |
| Shokati Asl et al. | Art and beauty in Rumi's mystical thought in accordance with Rudolf Otto and Maslow | |
| Margolis | Notes on Feyerabend and Hanson | |
| Pallara | The challenge of Open Source: The diffusion of knowledge VS the evaluation of the research | |
| KUCUKOGLU et al. | Some special numbers and polynomials related to k-ary Lyndon words | |
| Bais et al. | Simplifying the Definition of Ma'rifah | |
| SHERIGAR | OCULOMOTOR DEFICITS IN ATTENTION DEFICIT HYPERACTIVITY DISORDER (ADHD)-A SYSTEMATIC REVIEW AND META-ANALYSIS |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| BB1A | Laying open of patent application |
Effective date: 20210507 |
|
| FG3A | Patent granted, date of granting |
Effective date: 20220916 |