Global ETD Search
Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.
Results
Showing 1 to 20 of 50 for “"imperfect data"”.
-
ROBUST REPRESENTATION LEARNING FROM IMPERFECT DATA
… we explore practical solutions to the data imperfection problem and propose supporting theories for them. We present research on robust representation learning in three important fields, which are self-supervised learning from unlabeled data, domain adaptation for out-of-domain data, …
-
Representation learning with Less Label and Imperfect Data
… of annotated examples for training. However, the data annotating process is expensive, prone to noisy information and human errors, and time-consuming. Moreover, in many applications (such as in medical fields), this process requires domain knowledge and expertise, therefore, often unable to …
-
Automated Machine Learning with Constraints and Imperfect Data
… automatically. It could release the burden of data scientists from the multifarious manual tuning process. However, dataset does not always have correct labels and sufficient data size. Wrong labels disrupt the training procedure and could not provide representative evaluation performance for …
-
Dynamic State Estimation of Microgrid With Imperfect Data Communication
… accuracy of the state estimator. In addition to data loss or delay, sensor measurements may include outliers that distort state estimation. We utilized the Generalized Maximum Likelihood-extended Kalman filter (GM-EKF), as a robust estimator, which exhibits good tracking capabilities suppressing …
-
Stochastic Learning and Optimization with Imperfect Data in Cyber-Physical Systems
… of cyber-physical systems (CPSs) with imperfect data. CPSs are engineered systems that are built from, and depend upon, the seamless integration of computation and physical components. The recent success of data-driven approaches based on the collection of a large amount of data …
-
Advancing Digital Pathology with Interpretable AI: From Imperfect Data to Biological Insights
… challenges. From an input perspective, pathology data are high-dimensional and information-rich, whereas annotations are scarce, coarse, and noisy. When coupled with limited sample sizes, standard supervised learning methods may not be readily applicable. On the output side, many digital pathology …
-
Towards Automated Machine Learning on Imperfect Data for Situational Awareness in Power System
… devices in the electric grid, large amounts of data are being collected, which provide golden opportunities to achieve high level of situational awareness for reliable and cost-effective grid operations.To better utilize the data, this dissertation aims to develop Machine Learning (ML) methods …
-
Deep Learning for Lung Cancer Detection: An Analysis of the Effects of Imperfect Data and Model Biases
… publicly available and widely used Lung Image Database Consortium and Image Database Resource Initiative (LIDC-IDRI). The analysis delves deeper into the learnt latent representations to provide valuable insights about CNNs. The findings indicate the presence of biases with a strong …
-
Cost-Sensitive Learning-based Methods for Imbalanced Classification Problems with Applications
Analysis and predictive modeling of massive datasets is an extremely significant problem that arises in many practical applications. The task of predictive modeling becomes even more challenging when data are imperfect or uncertain. The real data are frequently affected by outliers, uncertain …
-
Requirements, specifications and deployment models for autonomous jobsite safety proximity monitoring
… accuracy of the specifications and systems under imperfect data. I further evaluate some approaches to dealing with imperfect data. Collectively, these advances move existing proximity warning research from evaluation of specific systems to an engineering approach to development and deployment of …
-
Productivity and measurement issues: An application to Brazilian agriculture
… examines measurement problems caused by imperfect data and productivity issues in a microeconometric framework. The empirical work is done on a panel of 384 farmers in the Zona da Mata region of Minas Gerais, Brazil, for the period extending from 1979-84. The farmers use modern, labor and …
-
A Comparison of the Quality of Rule Induction from Inconsistent Data Sets and Incomplete Data Sets
In data mining, decision rules induced from known examples are used to classify unseen cases. There are various rule induction algorithms, such as LEM1 (Learning from Examples Module version 1), LEM2 (Learning from Examples Module version 2) and MLEM2 (Modified Learning from Examples Module version …
-
Data analytics for real-time event identification using real phasor measurement unit data
… inherent physical laws of power systems from a data perspective. However, achieving real-time system event identification has been challenging due to the high granularity and non-stationary nature of PMU time series, imperfect data quality, and the expensive cost of collecting high-quality event …
-
An accuracy model for relational databases
<p>The relational database model has gained a great deal of attention over the past decades for its advantages; however, it suffers from a ubiquitous problem of data imperfection. Based on this situation, several new models have been proposed to solve these problems. These models could improve the …
-
Robust incremental relational learning
… Learning tasks can be complex and the training data noisy. When operating as part of a larger system, there may be limitations on available memory and computational resources. Learners may also be required to provide results from a stream. This thesis investigates the problem of incremental, …
-
Analysis of a long-range environmental forecasting model
… a five to twenty year range. On the basis of imperfect data that is available for these regions, this study examines the innovations introduced to handle the unstable situations found in developing areas of the world. Limited to the Middle East region, this effort undertakes a restructuring of …
-
DESIGNING TRUSTWORTHY AI UNDER REAL WORLD DATA CHALLENGES
… model size, computational power, and dataset volume to the greatest extent possible. Historically, AI systems have been found on the principle that a model is only as effective as the data on which it is trained. Although data can be carefully cleaned and annotated in controlled …
-
Integrating Deep Traffic Prediction and Environmental Impact Assessment Using Noisy Real-World Data in Las Vegas
… road network, drawing on spatial and temporal data from a large-scale sensor network. The study significantly expands the dataset from 26 to approximately 900 sensors, enhancing predictive accuracy and regional coverage. Inspired by masking techniques and strategies tailored to incomplete …
-
A Double-Edged Sword: The Role of Noise in Trustworthy Deep Learning
… important challenge: noise. Noise can come from imperfect data, human labeling errors, or low-quality sensors, and it can reduce the reliability of learning systems. At the same time, noise is deliberately added to protect people's privacy, especially when sensitive data such as facial images are …
-
Factor abundance and Korea's comparative advantage in international trade
… with the factor proportions theory; that, with imperfect data, the regression models provide a more consistent explanation than do other methods based on the measurement of the factor content of trade; and that, although the factor proportions theory is clearly important, other considerations …
Page 1 of 3