Matching Items (6)
Filtering by
- All Subjects: artificial intelligence
- All Subjects: Mechanical Engineering
Description
This dissertation transforms a set of system complexity reduction problems to feature selection problems. Three systems are considered: classification based on association rules, network structure learning, and time series classification. Furthermore, two variable importance measures are proposed to reduce the feature selection bias in tree models. Associative classifiers can achieve high accuracy, but the combination of many rules is difficult to interpret. Rule condition subset selection (RCSS) methods for associative classification are considered. RCSS aims to prune the rule conditions into a subset via feature selection. The subset then can be summarized into rule-based classifiers. Experiments show that classifiers after RCSS can substantially improve the classification interpretability without loss of accuracy. An ensemble feature selection method is proposed to learn Markov blankets for either discrete or continuous networks (without linear, Gaussian assumptions). The method is compared to a Bayesian local structure learning algorithm and to alternative feature selection methods in the causal structure learning problem. Feature selection is also used to enhance the interpretability of time series classification. Existing time series classification algorithms (such as nearest-neighbor with dynamic time warping measures) are accurate but difficult to interpret. This research leverages the time-ordering of the data to extract features, and generates an effective and efficient classifier referred to as a time series forest (TSF). The computational complexity of TSF is only linear in the length of time series, and interpretable features can be extracted. These features can be further reduced, and summarized for even better interpretability. Lastly, two variable importance measures are proposed to reduce the feature selection bias in tree-based ensemble models. It is well known that bias can occur when predictor attributes have different numbers of values. Two methods are proposed to solve the bias problem. One uses an out-of-bag sampling method called OOBForest, and the other, based on the new concept of a partial permutation test, is called a pForest. Experimental results show the existing methods are not always reliable for multi-valued predictors, while the proposed methods have advantages.
ContributorsDeng, Houtao (Author) / Runger, George C. (Thesis advisor) / Lohr, Sharon L (Committee member) / Pan, Rong (Committee member) / Zhang, Muhong (Committee member) / Arizona State University (Publisher)
Created2011
Description
Human fingertips contain thousands of specialized mechanoreceptors that enable effortless physical interactions with the environment. Haptic perception capabilities enable grasp and manipulation in the absence of visual feedback, as when reaching into one's pocket or wrapping a belt around oneself. Unfortunately, state-of-the-art artificial tactile sensors and processing algorithms are no match for their biological counterparts. Tactile sensors must not only meet stringent practical specifications for everyday use, but their signals must be processed and interpreted within hundreds of milliseconds. Control of artificial manipulators, ranging from prosthetic hands to bomb defusal robots, requires a constant reliance on visual feedback that is not entirely practical. To address this, we conducted three studies aimed at advancing artificial haptic intelligence. First, we developed a novel, robust, microfluidic tactile sensor skin capable of measuring normal forces on flat or curved surfaces, such as a fingertip. The sensor consists of microchannels in an elastomer filled with a liquid metal alloy. The fluid serves as both electrical interconnects and tunable capacitive sensing units, and enables functionality despite substantial deformation. The second study investigated the use of a commercially-available, multimodal tactile sensor (BioTac sensor, SynTouch) to characterize edge orientation with respect to a body fixed reference frame, such as a fingertip. Trained on data from a robot testbed, a support vector regression model was developed to relate haptic exploration actions to perception of edge orientation. The model performed comparably to humans for estimating edge orientation. Finally, the robot testbed was used to perceive small, finger-sized geometric features. The efficiency and accuracy of different haptic exploratory procedures and supervised learning models were assessed for estimating feature properties such as type (bump, pit), order of curvature (flat, conical, spherical), and size. This study highlights the importance of tactile sensing in situations where other modalities fail, such as when the finger itself blocks line of sight. Insights from this work could be used to advance tactile sensor technology and haptic intelligence for artificial manipulators that improve quality of life, such as prosthetic hands and wheelchair-mounted robotic hands.
ContributorsPonce Wong, Ruben Dario (Author) / Santos, Veronica J (Thesis advisor) / Artemiadis, Panagiotis K (Committee member) / Helms Tillery, Stephen I (Committee member) / Posner, Jonathan D (Committee member) / Runger, George C. (Committee member) / Arizona State University (Publisher)
Created2013
Description
Modern automotive and aerospace products are large cyber-physical system involving both software and hardware, composed of mechanical, electrical and electronic components. The increasing complexity of such systems is a major concern as it impacts development time and effort, as well as, initial and operational costs. Towards the goal of measuring complexity, the first step is to determine factors that contribute to it and metrics to qualify it. These complexity components can be further use to (a) estimate the cost of cyber-physical system, (b) develop methods that can reduce the cost of cyber-physical system and (c) make decision such as selecting one design from a set of possible solutions or variants. To determine the contributions to complexity we conducted survey at an aerospace company. We found out three types of contribution to the complexity of the system: Artifact complexity, Design process complexity and Manufacturing complexity. In all three domains, we found three types of metrics: size complexity, numeric complexity (degree of coupling) and technological complexity (solvability).We propose a formal representation for all three domains as graphs, but with different interpretations of entity (node) and relation (link) corresponding to the above three aspects. Complexities of these components are measured using algorithms defined in graph theory. Two experiments were conducted to check the meaningfulness and feasibility of the complexity metrics. First experiment was mechanical transmission and the scope of this experiment was component level. All the design stages, from concept to manufacturing, were considered in this experiment. The second experiment was conducted on hybrid powertrains. The scope of this experiment was assembly level and only artifact complexity is considered because of the limited resources. Finally the calibration of these complexity measures was conducted at an aerospace company but the results cannot be included in this thesis.
ContributorsGurpreet Singh (Author) / Shah, Jami (Thesis advisor) / Runger, George C. (Committee member) / Davidson, Joseph (Committee member) / Arizona State University (Publisher)
Created2011
Description
The Cognitive Decision Support (CDS) model is proposed. The model is widely applicable and scales to realistic, complex decision problems based on adaptive learning. The utility of a decision is discussed and four types of decisions associated with CDS model are identified. The CDS model is designed to learn decision utilities. Data enrichment is introduced to promote the effectiveness of learning. Grouping is introduced for large-scale decision learning. Introspection and adjustment are presented for adaptive learning. Triage recommendation is incorporated to indicate the trustworthiness of suggested decisions.
The CDS model and methodologies are integrated into an architecture using concepts from cognitive computing. The proposed architecture is implemented with an example use case to inventory management.
Reinforcement learning (RL) is discussed as an alternative, generalized adaptive learning engine for the CDS system to handle the complexity of many problems with unknown environments. An adaptive state dimension with context that can increase with newly available information is discussed. Several enhanced components for RL which are critical for complex use cases are integrated. Deep Q networks are embedded with the adaptive learning methodologies and applied to an example supply chain management problem on capacity planning.
A new approach using Ito stochastic processes is proposed as a more generalized method to generate non-stationary demands in various patterns that can be used in decision problems. The proposed method generates demands with varying non-stationary patterns, including trend, cyclical, seasonal, and irregular patterns. Conventional approaches are identified as special cases of the proposed method. Demands are illustrated in realistic settings for various decision models. Various statistical criteria are applied to filter the generated demands. The method is applied to a real-world example.
The CDS model and methodologies are integrated into an architecture using concepts from cognitive computing. The proposed architecture is implemented with an example use case to inventory management.
Reinforcement learning (RL) is discussed as an alternative, generalized adaptive learning engine for the CDS system to handle the complexity of many problems with unknown environments. An adaptive state dimension with context that can increase with newly available information is discussed. Several enhanced components for RL which are critical for complex use cases are integrated. Deep Q networks are embedded with the adaptive learning methodologies and applied to an example supply chain management problem on capacity planning.
A new approach using Ito stochastic processes is proposed as a more generalized method to generate non-stationary demands in various patterns that can be used in decision problems. The proposed method generates demands with varying non-stationary patterns, including trend, cyclical, seasonal, and irregular patterns. Conventional approaches are identified as special cases of the proposed method. Demands are illustrated in realistic settings for various decision models. Various statistical criteria are applied to filter the generated demands. The method is applied to a real-world example.
ContributorsKee, Seho (Author) / Runger, George C. (Thesis advisor) / Escobedo, Adolfo (Committee member) / Gel, Esma (Committee member) / Janakiram, Mani (Committee member) / Rogers, Dale (Committee member) / Arizona State University (Publisher)
Created2020
Description
All biological processes like cell growth, cell differentiation, development, and aging requires a series of steps which are characterized by gene regulation. Studies have shown that gene regulation is the key to various traits and diseases. Various factors affect the gene regulation which includes genetic signals, epigenetic tracks, genetic variants, etc. Deciphering and cataloging these functional genetic elements in the non-coding regions of the genome is one of the biggest challenges in precision medicine and genetic research. This thesis presents two different approaches to identifying these elements: TreeMap and DeepCORE. The first approach involves identifying putative causal genetic variants in cis-eQTL accounting for multisite effects and genetic linkage at a locus. TreeMap performs an organized search for individual and multiple causal variants using a tree guided nested machine learning method. DeepCORE on the other hand explores novel deep learning techniques that models the relationship between genetic, epigenetic and transcriptional patterns across tissues and cell lines and identifies co-operative regulatory elements that affect gene regulation. These two methods are believed to be the link for genotype-phenotype association and a necessary step to explaining various complex diseases and missing heritability.
ContributorsChandrashekar, Pramod Bharadwaj (Author) / Liu, Li (Thesis advisor) / Runger, George C. (Committee member) / Dinu, Valentin (Committee member) / Arizona State University (Publisher)
Created2020
Description
Feature learning and the discovery of nonlinear variation patterns in high-dimensional data is an important task in many problem domains, such as imaging, streaming data from sensors, and manufacturing. This dissertation presents several methods for learning and visualizing nonlinear variation in high-dimensional data. First, an automated method for discovering nonlinear variation patterns using deep learning autoencoders is proposed. The approach provides a functional mapping from a low-dimensional representation to the original spatially-dense data that is both interpretable and efficient with respect to preserving information. Experimental results indicate that deep learning autoencoders outperform manifold learning and principal component analysis in reproducing the original data from the learned variation sources.
A key issue in using autoencoders for nonlinear variation pattern discovery is to encourage the learning of solutions where each feature represents a unique variation source, which we define as distinct features. This problem of learning distinct features is also referred to as disentangling factors of variation in the representation learning literature. The remainder of this dissertation highlights and provides solutions for this important problem.
An alternating autoencoder training method is presented and a new measure motivated by orthogonal loadings in linear models is proposed to quantify feature distinctness in the nonlinear models. Simulated point cloud data and handwritten digit images illustrate that standard training methods for autoencoders consistently mix the true variation sources in the learned low-dimensional representation, whereas the alternating method produces solutions with more distinct patterns.
Finally, a new regularization method for learning distinct nonlinear features using autoencoders is proposed. Motivated in-part by the properties of linear solutions, a series of learning constraints are implemented via regularization penalties during stochastic gradient descent training. These include the orthogonality of tangent vectors to the manifold, the correlation between learned features, and the distributions of the learned features. This regularized learning approach yields low-dimensional representations which can be better interpreted and used to identify the true sources of variation impacting a high-dimensional feature space. Experimental results demonstrate the effectiveness of this method for nonlinear variation pattern discovery on both simulated and real data sets.
A key issue in using autoencoders for nonlinear variation pattern discovery is to encourage the learning of solutions where each feature represents a unique variation source, which we define as distinct features. This problem of learning distinct features is also referred to as disentangling factors of variation in the representation learning literature. The remainder of this dissertation highlights and provides solutions for this important problem.
An alternating autoencoder training method is presented and a new measure motivated by orthogonal loadings in linear models is proposed to quantify feature distinctness in the nonlinear models. Simulated point cloud data and handwritten digit images illustrate that standard training methods for autoencoders consistently mix the true variation sources in the learned low-dimensional representation, whereas the alternating method produces solutions with more distinct patterns.
Finally, a new regularization method for learning distinct nonlinear features using autoencoders is proposed. Motivated in-part by the properties of linear solutions, a series of learning constraints are implemented via regularization penalties during stochastic gradient descent training. These include the orthogonality of tangent vectors to the manifold, the correlation between learned features, and the distributions of the learned features. This regularized learning approach yields low-dimensional representations which can be better interpreted and used to identify the true sources of variation impacting a high-dimensional feature space. Experimental results demonstrate the effectiveness of this method for nonlinear variation pattern discovery on both simulated and real data sets.
ContributorsHoward, Phillip (Author) / Runger, George C. (Thesis advisor) / Montgomery, Douglas C. (Committee member) / Mirchandani, Pitu (Committee member) / Apley, Daniel (Committee member) / Arizona State University (Publisher)
Created2016