| 1. |
- Candefjord, Stefan, 1981, et al.
(författare)
-
Microwave technology for detecting traumatic intracranial bleedings: tests on phantom of subdural hematoma and numerical simulations
- 2017
-
Ingår i: Medical and Biological Engineering and Computing. - : Springer Science and Business Media LLC. - 1741-0444 .- 0140-0118. ; 55:8, s. 1177-1188
-
Tidskriftsartikel (refereegranskat)abstract
- Traumatic brain injury is the leading cause of death and severe disability for young people and a major public health problem for elderly. Many patients with intracranial bleeding are treated too late, because they initially show no symptoms of severe injury and are not transported to a trauma center. There is a need for a method to detect intracranial bleedings in the prehospital setting. In this study, we investigate whether broadband microwave technology (MWT) in conjunction with a diagnostic algorithm can detect subdural hematoma (SDH). A human cranium phantom and numerical simulations of SDH are used. Four phantoms with SDH 0, 40, 70 and 110 mL are measured with a MWT instrument. The simulated dataset consists of 1500 observations. Classification accuracy is assessed using fivefold cross-validation, and a validation dataset never used for training. The total accuracy is 100 and 82–96 % for phantom measurements and simulated data, respectively. Sensitivity and specificity for bleeding detection were 100 and 96 %, respectively, for the simulated data. SDH of different sizes is differentiated. The classifier requires training dataset size in order of 150 observations per class to achieve high accuracy. We conclude that the results indicate that MWT can detect and estimate the size of SDH. This is promising for developing MWT to be used for prehospital diagnosis of intracranial bleedings.
|
|
| 2. |
|
|
| 3. |
- Harrison, Josie, 1989, et al.
(författare)
-
Scalability in Building Component Data Annotation: Enhancing Facade Material Classification with Synthetic Data
- 2024
-
Ingår i: Proceedings of the European Conference on Computing in Construction. - 2684-1150.
-
Konferensbidrag (refereegranskat)abstract
- Computer vision models trained on Google Street View images can create material cadastres. However, current approaches need manually annotated datasets that are difficult to obtain and often have class imbalance. To address these challenges, this paper fine-tuned a Swin Transformer model on a synthetic dataset generated with OpenAI’s DALL E and compared the performance to a similar manually annotated dataset. Although manual annotation remains the gold standard, the synthetic dataset performance demonstrates a reasonable alternative. The findings will ease annotation needed to develop material cadastres, offering architects insights into opportunities for material reuse, thus contributing to the reduction of demolition waste.
|
|
| 4. |
- Somanath, Sanjay, 1994, et al.
(författare)
-
AI-baserad segementering av fasader för att optimera renovering i en större skala
- 2021
-
Ingår i: Bygg och teknik. - 0281-658X. ; 2021:2, s. 26-29
-
Tidskriftsartikel (övrigt vetenskapligt/konstnärligt)abstract
- Hur kan vi på ett automatiskt sätt skapa mer detaljerade 3D modeller av byggnader i digitala tvillingar och förbättra indata för att beräkna energibesparingspotentialer i befintliga byggnader? I en pilotstudie har vi undersökt hur maskininlärning kan användas för att extrahera information om fönstersättning och storlek i befintliga byggnader. Vi har utvecklat en modell som har “tränats” att känna igen och segmenterar fönster från bilder med byggnadsfasader och på så sätt skapa digitala och mer detaljerade data för befintliga byggnader. Vårt långsiktiga mål är att utveckla en helautomatisk metod för analyser av renoveringspotentialer för byggnader och fastighetsportföljer.
|
|
| 5. |
|
|
| 6. |
- Yu, Yinan, 1985, et al.
(författare)
-
climateBUG: A data-driven framework for analyzing bank reporting through a climate lens
- 2024
-
Ingår i: Expert Systems with Applications. - 0957-4174 .- 1873-6793. ; 239
-
Tidskriftsartikel (refereegranskat)abstract
- This paper applies computational linguistics learning methods to the banking industry and climate change fields. We introduce our data-driven framework, climateBUG, with the aim of detecting latent information about how banks discuss their activities related to climate change using natural language processing (NLP). This framework consists of an ingestion pipeline, a configurable database, and a set of API’s. In addition, climateBUG offers two standalone components, namely a unique annotated corpus of approximately 1.1M statements from EU banks’ annual and sustainability reporting and a deep learning model adapted to the semantics of the corpus. When benchmarking on classification performance, our model outperforms other models with similar scopes due to its stronger domain relevance. We also provide examples of how the framework can be applied from a user perspective.
|
|
| 7. |
- Yu, Yinan, 1985, et al.
(författare)
-
Qually: A Quality Validation Toolbox for Automotive Perception Data Towards Trustworthy AI
- 2022
-
Rapport (övrigt vetenskapligt/konstnärligt)abstract
- Data-driven techniques such as artificial intelligence (AI) and deep learning are frequently deployed as part of automotive perception systems. Due to their heavy dependency on data, data quality is at the essence. In particular, in an automotive perception system, data is captured by sensors and transformed into different formats depending on where it is in the AI data processing pipeline. Although data at different stages share similar attributes, the impact of their properties at each individual stage differ significantly from one another. Therefore, data quality requirements need to be defined specifically at each stage. In this project, the objective is to develop an end-to-end quality control toolbox to detect errors and anomalies throughout the entire pipeline. To achieve this objective, we divide the project into three work packages, where the first step is to design a set of data properties and their corresponding requirements as quality specifications for data at each stage. Given these specifications, as a second step, we have developed a toolbox, Qually, to evaluate data quality metrics and detect errors and anomalies throughout the AI pipeline. In the last work package, as a demonstrator, Qually is applied to improve automated annotations. This is implemented in three steps: 1) errors are identified using the quality metrics evaluated by Qually; 2) Qually suggests an automatic correction using ensemble techniques; 3) the corrected annotations are evaluated by Qually to confirm the improvement in quality. The error detection and suggested corrections are manually inspected to statistically validate the outcome of Qually. As the next step, besides further developing Qually as a software to improve its robustness, capacity, scalability and completeness, we plan to focus on enriching the set of data properties and quality specifications, especially by including technical and business requirements from various automotive stakeholders. We also plan to investigate the possibility and scalability of integrating formal verification techniques for quality control.
|
|
| 8. |
|
|
| 9. |
- Candefjord, Stefan, 1981, et al.
(författare)
-
Microwave technology for localization of traumatic intracranial bleedings—a numerical simulation study
- 2013
-
Ingår i: Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS. - 1557-170X. - 9781457702167 ; , s. 1948-1951
-
Konferensbidrag (refereegranskat)abstract
- Traumatic brain injury (TBI) is a major public health problem worldwide. Intracranial bleedings represents the most serious complication of TBI and need to be surgically evacuated promptly to save lives and mitigate injury. Microwave technology (MWT) is promising as a complement to computed tomography (CT) to be used in road and air ambulances for early detection of intracranial bleedings. In this study, we perform numerical simulations to investigate if a classification algorithm based on singular value decomposition can distinguish between bleedings at different positions adjacent to the skull bone for a similar but simplified problem. The classification accuracy is 94-100% for all classes, a result that encourages us to pursue our efforts with MWT for more realistic scenarios. This indicates that MWT has potential for localizing a detected bleeding, which would increase the diagnostic value of this technique.
|
|
| 10. |
|
|
| 11. |
- Dobsicek Trefna, Hana, 1979, et al.
(författare)
-
Detection of wood decay by microwaves
- 2015
-
Ingår i: Progress in Electromagnetics Research Symposium, PIERS 2015,Prague.. - 1559-9450.
-
Konferensbidrag (refereegranskat)
|
|
| 12. |
- Dona, Malsha Ashani Mahawatta, et al.
(författare)
-
AirDnD-Asynchronous In-Range Dynamic and Distributed Network Orchestration Framework
- 2023
-
Ingår i: Proceedings - International Conference on Distributed Computing Systems. ; 2023-July, s. 953-954
-
Konferensbidrag (refereegranskat)abstract
- The increasing usage of IoT devices has generated an extensive volume of data which resulted in the establishment of data centers with well-structured computing infrastructure. Reducing underutilized resources of such data centers can be achieved by monitoring the tasks and offloading them across various compute units. This approach can also be used in mini mobile data ponds generated by edge devices and smart vehicles. This research aims to improve and utilize the usage of computing resources in distributed edge devices by forming a dynamic mesh network. The nodes in the mesh network shall share their computing tasks with another node that possesses unused computing resources. This proposed method ensures the minimization of data transfer between entities. The proposed AirDnD vision will be applied to a practical scenario relevant to an autonomous vehicle that approaches an intersection commonly known as 'looking around the corner' in related literature, collecting essential computational results from nearby vehicles to enhance its perception. The proposed solution consists of three models that transform growing amounts of geographically distributed edge devices into a living organism.
|
|
| 13. |
- Fhager, Andreas, 1976, et al.
(författare)
-
Microwave Technology in Medical Diagnostics and Treatment
- 2015
-
Ingår i: 2015 Ieee Mtt-S International Microwave Workshop Series on Rf and Wireless Technologies for Biomedical and Healthcare Applications. - New York : Ieee. - 9781479985432 ; , s. 133-134
-
Bokkapitel (övrigt vetenskapligt/konstnärligt)abstract
- There is a great need for novel diagnostics and treatment tools in today's healthcare. In this paper we describe our development and progress in novel microwave based diagnostics and treatment applications. The target applications are stroke diagnostics, breast cancer detection and microwave hyperthermia.
|
|
| 14. |
- Fhager, Andreas, 1976, et al.
(författare)
-
Stroke diagnostics with a microwave helmet
- 2013
-
Ingår i: 2013 7th European Conference on Antennas and Propagation, EuCAP 2013. - 2164-3342. - 9788890701832 ; , s. 845-846
-
Konferensbidrag (refereegranskat)abstract
- In this paper we describe a microwave based measurement setup for detection and diagnostics of stroke. Results and experience from clinical testing is reported.
|
|
| 15. |
- Ho, Emily, et al.
(författare)
-
Sentiment and semantic analysis: Urban quality inference using machine learning algorithms
- 2024
-
Ingår i: iScience. - 2589-0042. ; 27:7
-
Tidskriftsartikel (refereegranskat)abstract
- Sustainable urban transformation requires comprehensive knowledge about the built environment, including people's perceptions, use of sites, and wishes. Qualitative interviews are conducted to understand better people's opinions about a specific topic or location. This study explores the automatization of the interview coding process by investigating how state-of-the-art natural language processing techniques classify sentiment and semantic orientation from interviews transcribed in Swedish. For the sentiment analysis, the Swedish bidirectional encoder representations from transformers (BERT) model KB-BERT was used to perform a multi-class classification task on a text sentence level into three different classes: positive, negative, and neutral. Named entity recognition (NER) and string search were used for the semantic analysis to perform multi-label classification to match domain-related topics to the sentence. The models were trained and evaluated on partially annotated datasets. The results demonstrate that the implemented deep learning techniques are a possible and promising solution to achieve the stated goal.
|
|
| 16. |
- Jian, Qiuchi, 1984, et al.
(författare)
-
Detection of breathing and heartbeat by using a simple UWB radar system
- 2014
-
Ingår i: 8th European Conference on Antennas and Propagation, EuCAP 2014. - 2164-3342. - 9788890701849 ; , s. 3078-3081
-
Konferensbidrag (refereegranskat)abstract
- We present the development on an ultra-wideband (UWB) radar system and its signal processing algorithms for detecting human breathing and heartbeat in the paper. The UWB radar system consists of two (Tx and Rx) antennas and one compact CMOS UWB transceiver. Several signal processing techniques are developed for the application. The system has been tested by real measurements. .
|
|
| 17. |
- Persson, Mikael, 1959, et al.
(författare)
-
Microwave based diagnostics and treatment in practice
- 2013
-
Ingår i: 2013 IEEE MTT-S International Microwave Workshop Series on RF and Wireless Technologies for Biomedical and Healthcare Applications, IMWS-BIO 2013 - Proceedings.
-
Konferensbidrag (refereegranskat)abstract
- Globally, around 15 million people each year suffer a stroke. Only a small fraction of stroke patients who could benefit from thrombolytic treatment reach diagnosis and treatment in time. To increase this low figure we have developed microwave technology aiming to differentiate hemorrhagic from ischemic stroke patients. The standard method for breast cancer diagnosis today is X-ray mammography. Despite its recognized ability to detect tumors it suffers from some limitations. Neither the false positive nor the false negative detection rates are negligible. An interesting alternative being researched extensively today is microwave tomography. In our current strive to develop a clinical prototype we have found that the most suitable design consists of an antenna array placed in a full 3D pattern. During the last decade clinical studies have demonstrated the ability of microwave hyperthermia to dramatically enhance cancer patient survival. The fundamental challenge is to adequately heat deep-seated tumors while preventing surrounding healthy tissue from undesired heating and damage. We are specifically addressing the challenge to deliver power levels with spatial control, patient treatment planning, and noninvasive temperature measurements. © 2013 IEEE.
|
|
| 18. |
- Persson, Mikael, 1959, et al.
(författare)
-
Microwave-Based Stroke Diagnosis Making Global Prehospital Thrombolytic Treatment Possible
- 2014
-
Ingår i: IEEE Transactions on Biomedical Engineering. - : Institute of Electrical and Electronics Engineers (IEEE). - 0018-9294 .- 1558-2531. ; 61:11, s. 2806-2817
-
Tidskriftsartikel (refereegranskat)abstract
- Here, we present two different brain diagnostic devices based on microwave technology and the associated two first proof-of-principle measurements that show that the systems can differentiate hemorrhagic from ischemic stroke in acute stroke patients, as well as differentiate hemorrhagic patients from healthy volunteers. The system was based on microwave scattering measurements with an antenna system worn on the head. Measurement data were analyzed with a machine-learning algorithm that is based on training using data from patients with a known condition. Computer tomography images were used as reference. The detection methodology was evaluated with the leave-one-out validation method combined with a Monte Carlo-based bootstrap step. The clinical motivation for this project is that ischemic stroke patients may receive acute thrombolytic treatment at hospitals, dramatically reducing or abolishing symptoms. A microwave system is suitable for prehospital use, and therefore has the potential to allow significantly earlier diagnosis and treatment than today.
|
|
| 19. |
- Persson, Mikael, 1959, et al.
(författare)
-
Pre-hospital care for stroke and trauma
- 2015
-
Ingår i: Conference Proceedings, 2014 IEEE MTT-S International Microwave Workshop Series on: RF and Wireless Technologies for Biomedical and Healthcare Applications, IMWS-Bio 2014, London, United Kingdom, 8-10 December 2014. - 9781479954476
-
Konferensbidrag (refereegranskat)abstract
- Pre-hospital care for stroke and trauma remains one of the global challenges., Each year around 15 million people each year suffer a stroke. Only a small fraction of stroke patients who could benefit from thrombolytic treatment reach diagnosis and treatment in time. To increase this low figure we have developed microwave technology aiming to differentiate hemorrhagic from ischemic stroke patients in a pre-hospital setting.
|
|
| 20. |
|
|
| 21. |
- von Numers, Charlotte, et al.
(författare)
-
BAUFER: A Baseline-Enabled Facial Expression Recognition Pipeline Trained With Limited Annotations
- 2023
-
Ingår i: Studies in Computational Intelligence, vol 1106. - 9783031369377
-
Bokkapitel (övrigt vetenskapligt/konstnärligt)abstract
- Social science theories suggest that facial expressions serve as a valuable indicator of one’s emotions, well-being, and overall functioning. Recent research has found that the facial expressions of participants in clinical trials can be linked to their self-reported quality of life. Since manual facial expression annotation and interpretation is time and cost intensive, automated facial expression recognition (FER) tools have the potential to make it quicker and more consistent to study an individual’s emotional responses. This paper introduces BAUFER, Baseline-enabled Action Unit identification for Facial Expression Recognition, with the following features: (1) a personalized baseline component to calibrate for the neutral expression of a participant; (2) predictions for anatomically-based facial muscle movement labels (Action Units), which have been reliably linked to emotional experiences in prior research, to enhance interpretability; and (3) a multi-stage training approach with several types of annotations from different datasets to overcome the known challenge of insufficient labeled data. While developed with non-clinical data, an intended future application of BAUFER is in the clinical domain to enhance our understanding of the patient experience.
|
|
| 22. |
- Yang, Jian, 1960, et al.
(författare)
-
Several new ultra-wideband antenna systems for radio telescopes and industry sensor imaging process
- 2012
-
Ingår i: 2012 14th International Conference on Electromagnetics in Advanced Applications, ICEAA 2012. Cape Town, 2 - 7 September 2012. - 9781467303354 ; , s. 1281-1284
-
Konferensbidrag (refereegranskat)abstract
- This paper presents an overview of several ultrawideband (UWB) antenna systems recently developed at Chalmers for applications in future UWB radio telescopes and industry sensor imaging process.
|
|
| 23. |
- Yu, Yinan, 1985, et al.
(författare)
-
A Classification Scheme for 'High-Dimensional-Small-Sample-Size' Data Using SODA and Ridge-SVM with Microwave Measurement Applications
- 2013
-
Ingår i: ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings. - 1520-6149. - 9781479903566 ; , s. 3542-3546
-
Konferensbidrag (refereegranskat)abstract
- The generalization performance of SVM-type classifiers severely suffers from the ‘curse of dimensionality’. For some real world applications, the dimensionality of the measurement is sometimes significantly larger compared to the amount of training data samples available. In this paper, a classification scheme is proposed and compared with existing techniques for such scenarios. The proposed scheme includes two parts: (i) feature selection and transformation based on Fisher discriminant criteria and (ii) a hybrid classifier combining Kernel Ridge Regression with Support Vector Machine to predict the label of the data. The first part is named Successively Orthogonal Discriminant Analysis (SODA), which is applied after Fisher score based feature selection as a preliminaryprocessing for dimensionality reduction. At this step, SODA maximizes the ratio of between-class-scatter and within-class-scatter to obtain an orthogonal transformation matrix which maps the features to a new low dimensional feature space where the class separability is maximized. The techniques are tested on high dimensional data from a microwave measurements system and are compared with existing techniques
|
|
| 24. |
- Yu, Yinan, 1985, et al.
(författare)
-
A new UWB radar system using UWB CMOS chip
- 2011
-
Ingår i: Proceedings of the 5th European Conference on Antennas and Propagation, EUCAP 2011. Rome, 11-15 April 2011. - 9788882020743 ; , s. 771-775
-
Konferensbidrag (refereegranskat)abstract
- A complete ultra-wideband (UWB) radar system is presented in this paper. The radar system consists of UWB CMOS radar transceiver and two compact directional UWB antennas for the purpose of transmission and reception. Different possible applications were investigated in the domain of ranging and tracking using this system along with fast and efficient signal processing algorithms. The results obtained are very promising for this new technology.
|
|
| 25. |
- Yu, Yinan, 1985, et al.
(författare)
-
A robust subspace classification scheme based on empirical intersection removal and sparse approximation
- 2015
-
Ingår i: Integrated Computer-Aided Engineering. - 1875-8835 .- 1069-2509. ; 22:1, s. 59-69
-
Tidskriftsartikel (refereegranskat)abstract
- Subspace models are widely used in many applications. By assuming an individual subspace model for each class, linear regression is applied and combined with minimum distance criteria for making the final decision. In a generalized subspace model, the full linear subspace of each class is split into subspaces with lower dimensions, and the unknown basis needs to be estimated with respect to the testing pattern using adaptively selected training samples. The training data selection is implemented using either least-squares regression or sparse approximation. In this paper, to further improve the classification performance, instead of attempting to minimize the regression error for each class, the between class separability is enhanced by a novel approach called Empirical Subspace Intersection (ESI) Removal technique. Evaluations are performed on (1) standard UCI data set, and (2) a computer aided system along with the proposed classification technique to determine the quality in wooden logs using microwave signals. The experimental results are shown and compared with classical methods.
|
|
| 26. |
- Yu, Yinan, 1985, et al.
(författare)
-
A Subspace Learning Algorithm For Microwave Scattering Signal Classification With Application To Wood Quality Assessment
- 2012
-
Ingår i: IEEE International Workshop on Machine Learning for Signal Processing, MLSP. - 2161-0371 .- 2161-0363. - 9781467310260
-
Konferensbidrag (refereegranskat)abstract
- A classification algorithm based on a linear subspace model has been developed and is presented in this paper. To further improve the classification results, the full linear subspace of each class is split into subspaces with lower dimensions and characterized by local coordinates constructed from automatically selected training data. The training data selection is implemented by optimizations with least squares constraints or L1 regularization. The working application is to determine the quality in wooden logs using microwave signals [1]. The experimental results are shown and compared with classical methods
|
|
| 27. |
- Yu, Yinan, 1985, et al.
(författare)
-
A Unified Subspace Classification Framework Developed for Diagnostic System Using Microwave Signal
- 2013
-
Ingår i: European Signal Processing Conference. - 2219-5491. - 9780992862602
-
Konferensbidrag (refereegranskat)abstract
- Subspace learning is widely used in many signal processing and statistical learning problems where the signal is assumably generated from a low dimensional space. In this paper, we present a unified classifier including several concepts from different subspace techniques, such as PCA, LRC, LDA, GLRT, etc. The objective is to project the original signal (usually of high dimension) into a smaller subspace with 1) within-class data structure preserved and 2) between-class-distance enhanced. A novel classification technique called Maximum Angle Subspace Classifier (MASC) is presented to achieve these purposes. To compensate for the computational complexity and non-convexity of MASC, an approximation is proposed as a trade-off between the classification performance and the computational issue. The approaches are applied to the problem of classifying high dimensional frequency measurements from a microwave based diagnostic system and results are compared with existing methods.
|
|
| 28. |
- Yu, Yinan, 1985, et al.
(författare)
-
Adaptive margin slack minimization in RKHS for classification
- 2016
-
Ingår i: ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings. - 1520-6149. - 9781479999880 ; 2016-May, s. 2319-2323
-
Konferensbidrag (refereegranskat)abstract
- In this paper, we design a novel regularized empirical risk minimization technique for classification called Adaptive Margin Slack Minimization (AMSM). The proposed method is based on minimizing a regularized upper bound of the misclassification error. Compared to the cost function of the classical L2-SVM, AMSM can be interpreted as minimizing a tighter bound with some additional flexibilities regarding the choice of marginal hyperplane. A hyperparameter-free adaptive algorithm is presented for finding a solution to the proposed risk function. Numerical results shows that AMSM outperforms L2-SVM on the tested standard datasets.
|
|
| 29. |
|
|
| 30. |
- Yu, Yinan, 1985, et al.
(författare)
-
CLAss-Specific Subspace Kernel Representations and Adaptive Margin Slack Minimization for Large Scale Classification
- 2018
-
Ingår i: IEEE Transactions on Neural Networks and Learning Systems. - 2162-237X .- 2162-2388. ; 29:2, s. 440 -456
-
Tidskriftsartikel (refereegranskat)abstract
- In kernel-based classification models, given limited computational power and storage capacity, operations over the full kernel matrix becomes prohibitive. In this paper, we propose a new supervised learning framework using kernel models for sequential data processing. The framework is based on two components that both aim at enhancing the classification capability with a subset selection scheme. The first part is a subspace projection technique in the reproducing kernel Hilbert space using a CLAss-specific Subspace Kernel representation for kernel approximation. In the second part, we propose a novel structural risk minimization algorithm called the adaptive margin slack minimization to iteratively improve the classification accuracy by an adaptive data selection. We motivate each part separately, and then integrate them into learning frameworks for large scale data. We propose two such frameworks: the memory efficient sequential processing for sequential data processing and the parallelized sequential processing for distributed computing with sequential data acquisition. We test our methods on several benchmark data sets and compared with the state-of-the-art techniques to verify the validity of the proposed techniques.
|
|
| 31. |
- Yu, Yinan, 1985
(författare)
-
Classification of High Dimensional Signals with Small Training Sample Size with Applications towards Microwave Based Detection Systems
- 2013
-
Licentiatavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Classification techniques attempt to resolve the problem of categorizing data into two or more classes. The data distribution is therefore the most critical fact to be aware of. Unfortunately, specifications of data generators are not available in real life and a probabilistic density parameterisation is not always applicable, especially for the situation of High Dimensional data with Low (training) Sample Size (HDLSS). This raises the importance of developing data driven techniques, where the data model is assumed according to partially accessible prior knowledge or cross-validation. There are various popular data assumptions, such as centroid-based models, linear subspace models, manifold data structures, etc, and one should take into consideration the model accuracy, computational complexity, generalization ability, and be aware of possibilities of overfitting. When the dimensionality of the data is much higher than the training sample size, all issues appear as its nature and there is no easy way to find a good trade-off.In this work, we mainly focus on the first two types of data models and develop corresponding classification techniques. The first objective is to automatically learn the data generating model with limited amount of training samples available. With the assumed data model, the second step is to maximize the class separability with respect to the model assumption. The applications studied encompass both simulated and measured microwave signals for stroke type diagnostics and wood quality assessment. The results are analyzed and compared with more classical approaches.
|
|
| 32. |
- Yu, Yinan, 1985, et al.
(författare)
-
Compact UWB Indoor and Through-Wall Radar with Precise Ranging and Tracking
- 2012
-
Ingår i: International Journal of Antennas and Propagation. - : Hindawi Limited. - 1687-5869 .- 1687-5877. ; 2012, s. Article ID 678590-
-
Tidskriftsartikel (refereegranskat)abstract
- Ultrawideband (UWB) technology has many advantages compared to its narrowbandcounterpart in many applications. We present a new compact low-cost UWB radar forindoor and through-wall scenario. The focus of the paper is on the development of the signalprocessing algorithms for ranging and tracking, taking into account the particular propertiesof the UWB CMOS transceiver and the radiation characteristics of the antennas. Theoreticalanalysis for the algorithms and their evaluations by measurements are presented in the paper. The ranging resolution of this UWB radar has achieved 1-2 mm RMS accuracy for a movingtarget in indoor environment over a short range, and Kalman tracking algorithm functionswell for the through-wall detection.
|
|
| 33. |
|
|
| 34. |
- Yu, Yinan, 1985, et al.
(författare)
-
Enhanced Distance Subset Approximation using Class-Specific Subspace Kernel Representation for Kernel Approximation
- 2016
-
Ingår i: IEEE International Workshop on Machine Learning for Signal Processing, MLSP. - 2161-0371 .- 2161-0363. - 9781509007462 ; 2016-November
-
Konferensbidrag (refereegranskat)abstract
- The computational complexity of kernel methods grows at least quadratically with respect to the training size and hence low rank kernel approximation techniques are commonly used. One of the most popular approximations is constructed by sub-sampling the training data. In this paper, we present a sampling algorithm called Enhanced Distance Subset Approximation (EDSA) based on a novel kernel function called CLAss-Specific Kernel (CLASK), which applies the idea of subspace clustering to low rank kernel approximation. By representing the kernel matrix based on a class-specific subspace model, it is allowed to use distinct kernel functions for different classes, which provides a better flexibility compared to classical kernel approximation techniques. Experimental results conducted on various UCI datasets are provided in order to verify the proposed techniques.
|
|
| 35. |
- Yu, Yinan, 1985, et al.
(författare)
-
Feature Reduction Based on Sum-of-SNR (SOSNR) Optimization
- 2014
-
Ingår i: ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings. - 1520-6149. - 9781479928927 ; , s. 6756-6760
-
Konferensbidrag (refereegranskat)abstract
- Dimensionality reduction plays an important role in machine learning techniques. In classification, data transformation aims to reduce the number of feature dimensions, whereas attempts to enhance the class separability. To this end, we propose a new classifier-independent criterion called 'Sum-of-Signal-to-Noise-Ratio' (SoSNR). A framework designed for maximization with respect to this criterion is presented and three types of algorithms, respectively based on (1) gradient, (2) deflation and (3) sparsity, are proposed. The techniques are conducted on standard UCI databases and compared to other related methods. Results show trade-offs between computational complexity and classification accuracy among different approaches.
|
|
| 36. |
|
|
| 37. |
- Yu, Yinan, 1985, et al.
(författare)
-
Kernel Subspace Learning for Pattern Classification
- 2018
-
Ingår i: Adaptive Learning Methods for Nonlinear System Modeling. ; , s. 127-147
-
Bokkapitel (övrigt vetenskapligt/konstnärligt)abstract
- Kernel methods are nonparametric feature extraction techniques that attempt to boost the learning capability of machine learning algorithms using nonlinear transformations. However, one major challenge in its basic form is that the computational complexity and the memory requirement do not scale well with respect to the training size. Kernel approximation is commonly employed to resolve this issue. Essentially, kernel approximation is equivalent to learning an approximated subspace in the high-dimensional feature vector space induced and characterized by the kernel function. With streaming data acquisition, approximated subspaces can be constructed adaptively. Explicit feature vectors are then extracted by a transformation onto the approximated subspace and linear learning techniques can be subsequently applied. From a computational point of view, operations in kernel methods can easily be parallelized and modern infrastructures can be utilized to achieve efficient computing. Moreover, the extracted explicit feature vectors can easily be interfaced with other learning techniques.
|
|
| 38. |
- Yu, Yinan, 1985
(författare)
-
Machine Learning Methods Using Class-specific Subspace Kernel Representations for Large-Scale Applications
- 2016
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Kernel techniques became popular due to and along with the rising success of Support Vector Machines (SVM). During the last two decades, the kernel idea itself has been extracted from SVM and is now widely studied as an independent subject. Essentially, kernel methods are nonlinear transformation techniques that take data from an input set to a high (possibly infinite) dimensional vector space, called the Reproducing Kernel Hilbert Space (RKHS), in which linear models can be applied. The original input set could be data from different domains and applications, such as tweets, ratings of movies, images, medical measurements, etc. The two spaces are connected by a Positive-Semi Definite (PSD) kernel function and all computations in the RKHS are evaluated on the low dimensional input set using the kernel function.Kernel methods are proven to be efficient on various applications. However, the computational complexity of most kernel algorithms typically grows cubically, or at least quadratically, with respect to the training size. This is due to the fact that a Gram kernel matrix needs to be constructed and/or inverted. To improve the scalability for large-scale training, kernel approximation techniques are employed, where the kernel matrix is assumed to have a low-rank structure. Essentially, this is equivalent to assuming a subspace model spanned by a subset of the training data in the RKHS. The task is hence to estimate the subspace with respect to some criteria, such as the reconstruction error, the discriminative power for classification tasks, etc.Based on these motivations, this thesis focuses on the development of scalable kernel techniques for supervised classification problems. Inspired by the idea of the subspace classifier and kernel clustering models, we have proposed the CLAss-specific Subspace Kernel (CLASK) representation, where class-specific kernel functions are applied and individual subspaces can be constructed accordingly. In this thesis work, an automatic model selection technique is proposed to choose the best multiple kernel functions for each class based on a criterion using the subspace projection distance. Moreover, subset selection and transformation techniques using CLASK are developed to further reduce the model complexity with an enhanced discriminative power for kernel approximation and classification. Furthermore, we have also proposed both a parallel and a sequential framework to tackle large-scale learning problems.
|
|
| 39. |
|
|
| 40. |
- Yu, Yinan, 1985, et al.
(författare)
-
Multiclass Ridge-adjusted Slack Variable Optimization Using Selected Basis for Fast Classification
- 2014
-
Ingår i: European Signal Processing Conference. - 2219-5491. - 9780992862619 ; , s. 1178-1182
-
Konferensbidrag (refereegranskat)abstract
- Kernel techniques for classification is especially challenging in terms of computation and memory requirement when data fall into more than two categories. In this paper, we extend a binary classification technique called Ridge-adjusted Slack Variable Optimization (RiSVO) to its multiclass counterpart where the label information encoding scheme allows the computational complexity to remain the same to the binary case. The main features of this technique are summarized as follows: (1) Only a subset of data are pre-selected to construct the basis for kernel computation; (2) Simultaneous active training set selection for all classes helps reduce complexity meanwhile improving robustness; (3) With the proposed active set selection criteria, inclusion property is verified empirically. Inclusion property means that once a pattern is excluded, it will no longer return to the active training set and therefore can be permanently removed from the training procedure. This property greatly reduce the complexity. The proposed techniques are evaluated on standard multiclass datasets MNIST, USPS, pendigits and letter which could be easily compared with existing results.
|
|
| 41. |
- Yu, Yinan, 1985, et al.
(författare)
-
Ridge-Adjusted Slack Variable Optimization for Supervised Classification
- 2013
-
Ingår i: IEEE International Workshop on Machine Learning for Signal Processing, MLSP. - 2161-0371 .- 2161-0363. - 9781479911806
-
Konferensbidrag (refereegranskat)abstract
- This paper presents an iterative classification algorithm called Ridge-adjusted Slack Variable Optimization (RiSVO). RiSVO is an iterative procedure with two steps: (1) A working subset of the training data is selected so as to reject "extreme" patterns. (2) the decision vector and threshold value are obtained by minimizing the energy function associated with the slack variables. From a computational perspective, we have established a sufficient condition for the "inclusion property" among successive working sets, which allows us to save computation time. Most importantly, under the inclusion property, the monotonic reduction of the energy function can be assured in both substeps at each iteration, thus assuring the convergence of the algorithm. Moreover, ridge regularization is incorporated to improve the robustness and better cope with over-fitting and ill-conditioned problems. To verify the proposed algorithm, we conducted simulations on three data sets from the UCI database: adult, shuttle and bank. Our simulation shows stability and convergence of the RiSVO method. The results also show improvement of performance over the SVM classifier.
|
|
