Deep learning is increasingly becoming a viable way of classifying all types of data. Modern deep learning algorithms, such as one dimensional convolutional neural networks, have demonstrated excellent performance in classifying time series data because of the ability to identify time invariant features. A primary challenge of deep learning for time series classification is the large amount of data required for training and many application domains, such as in medicine, have challenges obtaining sufficient data. Transfer learning is a deep learning method used to apply feature knowledge from one deep learning model to another; this is a powerful tool when both training datasets are similar and offers smaller datasets the power of more robust larger datasets. This makes it vital that the best source dataset is selected when performing transfer learning and presently there is no metric for this purpose. In this thesis a metric of predicting the performance of transfer learning is proposed. To develop this metric this research will focus on classification and transfer learning for human-activity-recognition time series data. For general time series data, finding temporal relations between signals is computationally intensive using non-deep learning techniques. Rather than time-series signal processing, a neural network autoencoder was used to first transform the source and target datasets into a time independent feature space. To compare and quantify the suitability of transfer learning datasets, two metrics were examined: i) average embedded signal from each dataset was used to calculate the distance between each datasets centroid, and ii) a Generative Adversarial Network (GAN) model was trained and the discriminator portion of the GAN is then used to assess the dissimilarity between source and target. This thesis measures a correlation between the distance between two dataset and their similarity, as well as the ability for a GAN to discriminate between two datasets and their similarity. The discriminator metric, however, does suffer from an upper limit of dissimilarity. These metrics were then used to predict the success of transfer learning from one dataset to another for the purpose of general time series classification. / Thesis / Master of Applied Science (MASc) / Over the past decade, advances in computational power and increases in data quantity have made deep learning a useful method of complex pattern recognition and classification in data. There is a growing desire to be able to use these complex algorithms on smaller quantities of data. To achieve this, a deep learning model is first trained on a larger dataset and then retrained on the smaller dataset; this is called transfer learning. For transfer learning to be effective, there needs to be a level of similarity between the two datasets so that properties from larger dataset can be learned and then refined using the smaller dataset. Therefore, it is of great interest to understand what level of similarity exists between the two datasets. The goal of this research is to provide a similarity metric between two time series classification datasets so that potential performance gains from transfer learning can be better understood. The measure of similarity between two time series datasets presents a unique challenge due to the nature of this data. To address this challenge an encoder approach was implemented to transform the time series data into a form where each signal example can be compared against one another. In this thesis, different similarity metrics were evaluated and correlated to the performance of a deep learning model allowing the prediction of how effective transfer learning may be when applied.

Autonomous vehicles are complex robotic and artificial intelligence systems working together to achieve safe operation in unstructured environments. The objective of this work is to provide a foundation to develop more advanced algorithms for off-road autonomy. The project explores the sensors used for off-road autonomy and the data capture process. Additionally, the point cloud data captured from lidar sensors is processed to restore some of the geometric information lost during sensor sampling. Because ground truth values are needed for quantitative comparison, the MAVS was leveraged to generate a large off-road dataset in a variety of ecosystems. The results demonstrate data capture from the sensor suite and successful reconstruction of the selected geometric information. Using this geometric information, the point cloud data is more accurately segmented using the SqueezeSeg network.

As a complicated lethal medical emergency, sepsis is not easy to be diagnosed until it is too late for taking any life saving actions. Early prediction of sepsis in ICUs may reduce inpatient mortality rate. Although deep learning models can make predictions on the outcome of ICU stays with high accuracies, the opacity of such neural networks decreases their reliability. Particularly, in the ICU settings where the time is not on doctors' side and every single mistake increase the chances of patient's mortality. Therefore, it is crucial for the predictive model to provide some sort of reasoning in addition to the prediction it provides, so that the medical staff could avoid actions based on false alarms. To address this problem, we propose to add an attention layer to a deep recurrent neural network that can learn the relative importance of each of the parameters of the multivariate data of the ICU stay. Our approach sheds light on providing explainability through attention mechanism. We compare our method with some of the state-of-the-art methods and show the superiority of our approach in terms of providing explanations.

No description available.

No description available.

<p dir="ltr">The electronic health record (EHR) documents the patient’s medical history, with information such as demographics, diagnostic history, procedures, laboratory tests, and observations made by healthcare providers. This source of information can help support preventive health care and management. The present thesis explores the potential for EHR-driven models to predict acute care utilization (ACU) which is defined as visits to an emergency department (ED) or inpatient hospitalization (IH). ACU care is often associated with significant costs compared to outpatient visits. Identifying patients at risk can improve the quality of care for patients and can reduce the need for these services making healthcare organizations more cost-effective. This is important for vulnerable patients including those suffering from schizophrenia and bipolar disorders. This study compares the ability of the MedBERT architecture, the MedBERT+ architecture and standard machine learning models to identify at risk patients. MedBERT is a deep learning language model which was trained on diagnosis codes to predict the patient’s at risk for certain disease conditions. MedBERT+, the architecture introduced in this study is also trained on diagnosis codes. However, it adds socio-demographic embeddings and targets a different outcome, namely ACU. MedBERT+ outperformed the original architecture, MedBERT, as well as XGB achieving an AUC of 0.71 for both bipolar and schizophrenia patients when predicting ED visits and an AUC of 0.72 for bipolar patients when predicting IH visits. For schizophrenia patients, the IH predictive model had an AUC of 0.66 requiring further improvements. One potential direction for future improvement is the encoding of the demographic variables. Preliminary results indicate that an appropriate encoding of the age of the patient increased the AUC of Bipolar ED models to up to 0.78.</p>

京都大学 / 新制・課程博士 / 博士(エネルギー科学) / 甲第24009号 / エネ博第445号 / 新制||エネ||84(附属図書館) / 京都大学大学院エネルギー科学研究科エネルギー応用科学専攻 / (主査)教授 馬渕 守, 教授 土井 俊哉, 教授 濵 孝之 / 学位規則第4条第1項該当 / Doctor of Energy Science / Kyoto University / DGAM

京都大学 / 新制・課程博士 / 博士(情報学) / 甲第24034号 / 情博第790号 / 新制||情||134(附属図書館) / 京都大学大学院情報学研究科社会情報学専攻 / (主査)教授 黒田 知宏, 教授 吉川 正俊, 教授 神田 崇行 / 学位規則第4条第1項該当 / Doctor of Informatics / Kyoto University / DFAM

<p>Though modern deep learning based approaches have achieved remarkable progress in computer vision community such as image classification using a static image dataset, it suf- fers from catastrophic forgetting when learning new classes incrementally in a phase-by-phase fashion, in which only data for new classes are provided at each learning phase. In this work we focus on continual learning with the objective of learning new tasks from sequentially available data without forgetting the learned knowledge. We study this problem from three perspectives including (1) continual learning in online scenario where each data is used only once for training (2) continual learning in unsupervised scenario where no class label is pro- vided and (3) continual learning in real world applications. Specifically, for problem (1), we proposed a variant of knowledge distillation loss together with a two-step learning technique to efficiently maintain the learned knowledge and a novel candidates selection algorithm to reduce the prediction bias towards new classes. For problem (2), we introduced a new framework for unsupervised continual learning by using pseudo labels obtained from cluster assignments and an efficient out-of-distribution detector is designed to identify whether each new data belongs to new or learned classes in unsupervised scenario. For problem (3), we proposed a novel training regime targeted on food images using balanced training batch and a more efficient exemplar selection algorithm. Besides, we further proposed an exemplar-free continual learning approach to address the memory issue and privacy concerns caused by storing part of old data as exemplars.</p> <p>In addition to the work related to continual learning, we study the image-based dietary assessment with the objective of determining what someone eats and how much energy is consumed during the course of a day by using food or eating scene images. Specifically, we proposed a multi-task framework for simultaneously classification and portion size estima- tion by future fusion and soft-parameter sharing between backbone networks. Besides, we introduce RGB-Distribution image by concatenating the RGB image with the energy distri- bution map as the fourth channel, which is then used for end-to-end multi-food recognition and portion size estimation.</p>

Facade parsing is a fundamental problem in urban modeling that forms the back- bone of a variety of tasks including procedural modeling, architectural analysis, urban reconstruction and quite often relies on semantic segmentation as the first step. With the shift to deep learning based approaches, existing small-scale datasets are the bot- tleneck for making further progress in fa ̧cade segmentation and consequently fa ̧cade parsing. In this thesis, we propose a new fa ̧cade image dataset for semantic segmenta- tion called PSV-22, which is the largest such dataset. We show that PSV-22 captures semantics of fa ̧cades better than existing datasets. Additionally, we propose three architectural modifications to current state of the art deep-learning based semantic segmentation architectures and show that these modifications improve performance on our dataset and already existing datasets. Our modifications are generalizable to a large variety of semantic segmentation nets, but are fa ̧cade-specific and employ heuris- tics which arise from the regular grid-like nature of fac ̧ades. Furthermore, results show that our proposed architecture modifications improve the performance compared to baseline models as well as specialized segmentation approaches on fa ̧cade datasets and are either close in, or improve performance on existing datasets. We show that deep models trained on existing data have a substantial performance reduction on our data, whereas models trained only on our data actually improve when evaluated on existing datasets. We intend to release the dataset publically in the future.