Matching Items (112)
Filtering by
- All Subjects: artificial intelligence
- Genre: Masters Thesis
- Resource Type: Text
Description
For this master's thesis, an open learner model is integrated with Quinn, a teachable robotic agent developed at Arizona State University. This system is represented as a feedback system, which aims to improve a student’s understanding of a subject. It also helps to understand the effect of the learner model when it is represented by performance of the teachable agent. The feedback system represents performance of the teachable agent, and not of a student. Data in the feedback system is thus updated according to a student's understanding of the subject. This provides students an opportunity to enhance their understanding of a subject by analyzing their performance. To test the effectiveness of the feedback system, student understanding in two different conditions is analyzed. In the first condition a feedback report is not provided to the students, while in the second condition the feedback report is provided in the form of the agent’s performance.
ContributorsUpadhyay, Abha (Author) / Walker, Erin (Thesis advisor) / Nelson, Brian (Committee member) / Amresh, Ashish (Committee member) / Arizona State University (Publisher)
Created2016
Description
Speech recognition and keyword detection are becoming increasingly popular applications for mobile systems. While deep neural network (DNN) implementation of these systems have very good performance,
they have large memory and compute resource requirements, making their implementation on a mobile device quite challenging. In this thesis, techniques to reduce the memory and computation cost
of keyword detection and speech recognition networks (or DNNs) are presented.
The first technique is based on representing all weights and biases by a small number of bits and mapping all nodal computations into fixed-point ones with minimal degradation in the
accuracy. Experiments conducted on the Resource Management (RM) database show that for the keyword detection neural network, representing the weights by 5 bits results in a 6 fold reduction in memory compared to a floating point implementation with very little loss in performance. Similarly, for the speech recognition neural network, representing the weights by 6 bits results in a 5 fold reduction in memory while maintaining an error rate similar to a floating point implementation. Additional reduction in memory is achieved by a technique called weight pruning,
where the weights are classified as sensitive and insensitive and the sensitive weights are represented with higher precision. A combination of these two techniques helps reduce the memory
footprint by 81 - 84% for speech recognition and keyword detection networks respectively.
Further reduction in memory size is achieved by judiciously dropping connections for large blocks of weights. The corresponding technique, termed coarse-grain sparsification, introduces
hardware-aware sparsity during DNN training, which leads to efficient weight memory compression and significant reduction in the number of computations during classification without
loss of accuracy. Keyword detection and speech recognition DNNs trained with 75% of the weights dropped and classified with 5-6 bit weight precision effectively reduced the weight memory
requirement by ~95% compared to a fully-connected network with double precision, while showing similar performance in keyword detection accuracy and word error rate.
they have large memory and compute resource requirements, making their implementation on a mobile device quite challenging. In this thesis, techniques to reduce the memory and computation cost
of keyword detection and speech recognition networks (or DNNs) are presented.
The first technique is based on representing all weights and biases by a small number of bits and mapping all nodal computations into fixed-point ones with minimal degradation in the
accuracy. Experiments conducted on the Resource Management (RM) database show that for the keyword detection neural network, representing the weights by 5 bits results in a 6 fold reduction in memory compared to a floating point implementation with very little loss in performance. Similarly, for the speech recognition neural network, representing the weights by 6 bits results in a 5 fold reduction in memory while maintaining an error rate similar to a floating point implementation. Additional reduction in memory is achieved by a technique called weight pruning,
where the weights are classified as sensitive and insensitive and the sensitive weights are represented with higher precision. A combination of these two techniques helps reduce the memory
footprint by 81 - 84% for speech recognition and keyword detection networks respectively.
Further reduction in memory size is achieved by judiciously dropping connections for large blocks of weights. The corresponding technique, termed coarse-grain sparsification, introduces
hardware-aware sparsity during DNN training, which leads to efficient weight memory compression and significant reduction in the number of computations during classification without
loss of accuracy. Keyword detection and speech recognition DNNs trained with 75% of the weights dropped and classified with 5-6 bit weight precision effectively reduced the weight memory
requirement by ~95% compared to a fully-connected network with double precision, while showing similar performance in keyword detection accuracy and word error rate.
ContributorsArunachalam, Sairam (Author) / Chakrabarti, Chaitali (Thesis advisor) / Seo, Jae-Sun (Thesis advisor) / Cao, Yu (Committee member) / Arizona State University (Publisher)
Created2016
Description
For the past three decades, the design of an effective strategy for generating poetry that matches that of a human’s creative capabilities and complexities has been an elusive goal in artificial intelligence (AI) and natural language generation (NLG) research, and among linguistic creativity researchers in particular. This thesis presents a novel approach to fixed verse poetry generation using neural word embeddings. During the course of generation, a two layered poetry classifier is developed. The first layer uses a lexicon based method to classify poems into types based on form and structure, and the second layer uses a supervised classification method to classify poems into subtypes based on content with an accuracy of 92%. The system then uses a two-layer neural network to generate poetry based on word similarities and word movements in a 50-dimensional vector space.
The verses generated by the system are evaluated using rhyme, rhythm, syllable counts and stress patterns. These computational features of language are considered for generating haikus, limericks and iambic pentameter verses. The generated poems are evaluated using a Turing test on both experts and non-experts. The user study finds that only 38% computer generated poems were correctly identified by nonexperts while 65% of the computer generated poems were correctly identified by experts. Although the system does not pass the Turing test, the results from the Turing test suggest an improvement of over 17% when compared to previous methods which use Turing tests to evaluate poetry generators.
The verses generated by the system are evaluated using rhyme, rhythm, syllable counts and stress patterns. These computational features of language are considered for generating haikus, limericks and iambic pentameter verses. The generated poems are evaluated using a Turing test on both experts and non-experts. The user study finds that only 38% computer generated poems were correctly identified by nonexperts while 65% of the computer generated poems were correctly identified by experts. Although the system does not pass the Turing test, the results from the Turing test suggest an improvement of over 17% when compared to previous methods which use Turing tests to evaluate poetry generators.
ContributorsMagge, Arjun (Author) / Syrotiuk, Violet R. (Thesis advisor) / Baral, Chitta (Committee member) / Hogue, Cynthia (Committee member) / Bazzi, Rida (Committee member) / Arizona State University (Publisher)
Created2016
Description
The human motion is defined as an amalgamation of several physical traits such as bipedal locomotion, posture and manual dexterity, and mental expectation. In addition to the “positive” body form defined by these traits, casting light on the body produces a “negative” of the body: its shadow. We often interchangeably use with silhouettes in the place of shadow to emphasize indifference to interior features. In a manner of speaking, the shadow is an alter ego that imitates the individual.
The principal value of shadow is its non-invasive behaviour of reflecting precisely the actions of the individual it is attached to. Nonetheless we can still think of the body’s shadow not as the body but its alter ego.
Based on this premise, my thesis creates an experiential system that extracts the data related to the contour of your human shape and gives it a texture and life of its own, so as to emulate your movements and postures, and to be your extension. In technical terms, my thesis extracts abstraction from a pre-indexed database that could be generated from an offline data set or in real time to complement these actions of a user in front of a low-cost optical motion capture device like the Microsoft Kinect. This notion could be the system’s interpretation of the action which creates modularized art through the abstraction’s ‘similarity’ to the live action.
Through my research, I have developed a stable system that tackles various connotations associated with shadows and the need to determine the ideal features that contribute to the relevance of the actions performed. The implication of Factor Oracle [3] pattern interpretation is tested with a feature bin of videos. The system also is flexible towards several methods of Nearest Neighbours searches and a machine learning module to derive the same output. The overall purpose is to establish this in real time and provide a constant feedback to the user. This can be expanded to handle larger dynamic data.
In addition to estimating human actions, my thesis best tries to test various Nearest Neighbour search methods in real time depending upon the data stream. This provides a basis to understand varying parameters that complement human activity recognition and feature matching in real time.
The principal value of shadow is its non-invasive behaviour of reflecting precisely the actions of the individual it is attached to. Nonetheless we can still think of the body’s shadow not as the body but its alter ego.
Based on this premise, my thesis creates an experiential system that extracts the data related to the contour of your human shape and gives it a texture and life of its own, so as to emulate your movements and postures, and to be your extension. In technical terms, my thesis extracts abstraction from a pre-indexed database that could be generated from an offline data set or in real time to complement these actions of a user in front of a low-cost optical motion capture device like the Microsoft Kinect. This notion could be the system’s interpretation of the action which creates modularized art through the abstraction’s ‘similarity’ to the live action.
Through my research, I have developed a stable system that tackles various connotations associated with shadows and the need to determine the ideal features that contribute to the relevance of the actions performed. The implication of Factor Oracle [3] pattern interpretation is tested with a feature bin of videos. The system also is flexible towards several methods of Nearest Neighbours searches and a machine learning module to derive the same output. The overall purpose is to establish this in real time and provide a constant feedback to the user. This can be expanded to handle larger dynamic data.
In addition to estimating human actions, my thesis best tries to test various Nearest Neighbour search methods in real time depending upon the data stream. This provides a basis to understand varying parameters that complement human activity recognition and feature matching in real time.
ContributorsSeshasayee, Sudarshan Prashanth (Author) / Sha, Xin Wei (Thesis advisor) / Turaga, Pavan (Thesis advisor) / Tinapple, David A (Committee member) / Arizona State University (Publisher)
Created2016
Description
Detecting cyber-attacks in cyber systems is essential for protecting cyber infrastructures from cyber-attacks. It is very difficult to detect cyber-attacks in cyber systems due to their high complexity. The accuracy of the attack detection in the cyber systems depends heavily on the completeness of the collected sensor information. In this thesis, two approaches are presented: one to detecting attacks in completely observable cyber systems, and the other to estimating types of states in partially observable cyber systems for attack detection in cyber systems. These two approaches are illustrated using three large data sets of network traffic because the packet-level information of the network traffic data provides details about the cyber systems.
The approach to attack detection in cyber systems is based on a multimodal artificial neural network (MANN) using the collected network traffic data from completely observable cyber systems for training and testing. Since the training of MANN is computationally intensive, to reduce the computational overhead, an efficient feature selection algorithm using the genetic algorithm is developed and incorporated in this approach.
In order to detect attacks in cyber systems in partially observable environments, an approach to estimating the types of states in partially observable cyber systems, which is the first phase of attack detection in cyber systems in partially observable environments, is presented. The types of states of such cyber systems are useful to detecting cyber-attacks in such cyber systems. This approach involves the use of a convolutional neural network (CNN), and unsupervised learning with elbow method and k-means clustering algorithm.
The approach to attack detection in cyber systems is based on a multimodal artificial neural network (MANN) using the collected network traffic data from completely observable cyber systems for training and testing. Since the training of MANN is computationally intensive, to reduce the computational overhead, an efficient feature selection algorithm using the genetic algorithm is developed and incorporated in this approach.
In order to detect attacks in cyber systems in partially observable environments, an approach to estimating the types of states in partially observable cyber systems, which is the first phase of attack detection in cyber systems in partially observable environments, is presented. The types of states of such cyber systems are useful to detecting cyber-attacks in such cyber systems. This approach involves the use of a convolutional neural network (CNN), and unsupervised learning with elbow method and k-means clustering algorithm.
ContributorsGuha, Sayantan (Author) / Yau, Stephen S. (Thesis advisor) / Ahn, Gail-Joon (Committee member) / Huang, Dijiang (Committee member) / Arizona State University (Publisher)
Created2016
Description
Computational visual aesthetics has recently become an active research area. Existing state-of-art methods formulate this as a binary classification task where a given image is predicted to be beautiful or not. In many applications such as image retrieval and enhancement, it is more important to rank images based on their aesthetic quality instead of binary-categorizing them. Furthermore, in such applications, it may be possible that all images belong to the same category. Hence determining the aesthetic ranking of the images is more appropriate. To this end, a novel problem of ranking images with respect to their aesthetic quality is formulated in this work. A new data-set of image pairs with relative labels is constructed by carefully selecting images from the popular AVA data-set. Unlike in aesthetics classification, there is no single threshold which would determine the ranking order of the images across the entire data-set.
This problem is attempted using a deep neural network based approach that is trained on image pairs by incorporating principles from relative learning. Results show that such relative training procedure allows the network to rank the images with a higher accuracy than a state-of-art network trained on the same set of images using binary labels. Further analyzing the results show that training a model using the image pairs learnt better aesthetic features than training on same number of individual binary labelled images.
Additionally, an attempt is made at enhancing the performance of the system by incorporating saliency related information. Given an image, humans might fixate their vision on particular parts of the image, which they might be subconsciously intrigued to. I therefore tried to utilize the saliency information both stand-alone as well as in combination with the global and local aesthetic features by performing two separate sets of experiments. In both the cases, a standard saliency model is chosen and the generated saliency maps are convoluted with the images prior to passing them to the network, thus giving higher importance to the salient regions as compared to the remaining. Thus generated saliency-images are either used independently or along with the global and the local features to train the network. Empirical results show that the saliency related aesthetic features might already be learnt by the network as a sub-set of the global features from automatic feature extraction, thus proving the redundancy of the additional saliency module.
This problem is attempted using a deep neural network based approach that is trained on image pairs by incorporating principles from relative learning. Results show that such relative training procedure allows the network to rank the images with a higher accuracy than a state-of-art network trained on the same set of images using binary labels. Further analyzing the results show that training a model using the image pairs learnt better aesthetic features than training on same number of individual binary labelled images.
Additionally, an attempt is made at enhancing the performance of the system by incorporating saliency related information. Given an image, humans might fixate their vision on particular parts of the image, which they might be subconsciously intrigued to. I therefore tried to utilize the saliency information both stand-alone as well as in combination with the global and local aesthetic features by performing two separate sets of experiments. In both the cases, a standard saliency model is chosen and the generated saliency maps are convoluted with the images prior to passing them to the network, thus giving higher importance to the salient regions as compared to the remaining. Thus generated saliency-images are either used independently or along with the global and the local features to train the network. Empirical results show that the saliency related aesthetic features might already be learnt by the network as a sub-set of the global features from automatic feature extraction, thus proving the redundancy of the additional saliency module.
ContributorsGattupalli, Jaya Vijetha (Author) / Li, Baoxin (Thesis advisor) / Davulcu, Hasan (Committee member) / Liang, Jianming (Committee member) / Arizona State University (Publisher)
Created2016
Description
Sports activities have been a cornerstone in the evolution of humankind through the ages from the ancient Roman empire to the Olympics in the 21st century. These activities have been used as a benchmark to evaluate the how humans have progressed through the sands of time. In the 21st century, machines along with the help of powerful computing and relatively new computing paradigms have made a good case for taking up the mantle. Even though machines have been able to perform complex tasks and maneuvers, they have struggled to match the dexterity, coordination, manipulability and acuteness displayed by humans. Bi-manual tasks are more complex and bring in additional variables like coordination into the task making it harder to evaluate.
A task capable of demonstrating the above skillset would be a good measure of the progress in the field of robotic technology. Therefore a dual armed robot has been built and taught to handle the ball and make the basket successfully thus demonstrating the capability of using both arms. A combination of machine learning techniques, Reinforcement learning, and Imitation learning has been used along with advanced optimization algorithms to accomplish the task.
A task capable of demonstrating the above skillset would be a good measure of the progress in the field of robotic technology. Therefore a dual armed robot has been built and taught to handle the ball and make the basket successfully thus demonstrating the capability of using both arms. A combination of machine learning techniques, Reinforcement learning, and Imitation learning has been used along with advanced optimization algorithms to accomplish the task.
ContributorsKalige, Nikhil (Author) / Amor, Heni Ben (Thesis advisor) / Shrivastava, Aviral (Committee member) / Zhang, Yu (Committee member) / Arizona State University (Publisher)
Created2016
Description
Automated planning addresses the problem of generating a sequence of actions that enable a set of agents to achieve their goals.This work investigates two important topics from the field of automated planning, namely model-lite planning and multi-agent planning. For model-lite planning, I focus on a prominent model named Annotated PDDL and it's related application of robust planning. For this model, I try to identify a method of leveraging additional domain information (available in the form of successful plan traces). I use this information to refine the set of possible domains to generate more robust plans (as compared to the original planner) for any given problem. This method also provides us a way of overcoming one of the major drawbacks of the original approach, namely the need for a domain writer to explicitly identify the annotations.
For the second topic, the central question I ask is ``{\em under what conditions are multiple agents actually needed to solve a given planning problem?}''. To answer this question, the multi-agent planning (MAP) problem is classified into several sub-classes and I identify the conditions in each of these sub-classes that can lead to required cooperation (RC). I also identify certain sub-classes of multi-agent planning problems (named DVC-RC problems), where the problems can be simplified using a single virtual agent. This insight is later used to propose a new planner designed to solve problems from these subclasses. Evaluation of this new planner on all the current multi-agent planning benchmarks reveals that most current multi-agent planning benchmarks only belong to a small subset of possible classes of multi-agent planning problems.
For the second topic, the central question I ask is ``{\em under what conditions are multiple agents actually needed to solve a given planning problem?}''. To answer this question, the multi-agent planning (MAP) problem is classified into several sub-classes and I identify the conditions in each of these sub-classes that can lead to required cooperation (RC). I also identify certain sub-classes of multi-agent planning problems (named DVC-RC problems), where the problems can be simplified using a single virtual agent. This insight is later used to propose a new planner designed to solve problems from these subclasses. Evaluation of this new planner on all the current multi-agent planning benchmarks reveals that most current multi-agent planning benchmarks only belong to a small subset of possible classes of multi-agent planning problems.
ContributorsSreedharan, Sarath (Author) / Kambhampati, Subbarao (Thesis advisor) / Zhang, Yu (Thesis advisor) / Ben Amor, Heni (Committee member) / Arizona State University (Publisher)
Created2016
Description
To ensure system integrity, robots need to proactively avoid any unwanted physical perturbation that may cause damage to the underlying hardware. In this thesis work, we investigate a machine learning approach that allows robots to anticipate impending physical perturbations from perceptual cues. In contrast to other approaches that require knowledge about sources of perturbation to be encoded before deployment, our method is based on experiential learning. Robots learn to associate visual cues with subsequent physical perturbations and contacts. In turn, these extracted visual cues are then used to predict potential future perturbations acting on the robot. To this end, we introduce a novel deep network architecture which combines multiple sub- networks for dealing with robot dynamics and perceptual input from the environment. We present a self-supervised approach for training the system that does not require any labeling of training data. Extensive experiments in a human-robot interaction task show that a robot can learn to predict physical contact by a human interaction partner without any prior information or labeling. Furthermore, the network is able to successfully predict physical contact from either depth stream input or traditional video input or using both modalities as input.
ContributorsSur, Indranil (Author) / Amor, Heni B (Thesis advisor) / Fainekos, Georgios (Committee member) / Yang, Yezhou (Committee member) / Arizona State University (Publisher)
Created2017
Description
Ensemble learning methods like bagging, boosting, adaptive boosting, stacking have traditionally shown promising results in improving the predictive accuracy in classification. These techniques have recently been widely used in various domains and applications owing to the improvements in computational efficiency and distributed computing advances. However, with the advent of wide variety of applications of machine learning techniques to class imbalance problems, further focus is needed to evaluate, improve and optimize other performance measures such as sensitivity (true positive rate) and specificity (true negative rate) in classification. This thesis demonstrates a novel approach to evaluate and optimize the performance measures (specifically sensitivity and specificity) using ensemble learning methods for classification that can be especially useful in class imbalanced datasets. In this thesis, ensemble learning methods (specifically bagging and boosting) are used to optimize the performance measures (sensitivity and specificity) on a UC Irvine (UCI) 130 hospital diabetes dataset to predict if a patient will be readmitted to the hospital based on various feature vectors. From the experiments conducted, it can be empirically concluded that, by using ensemble learning methods, although accuracy does improve to some margin, both sensitivity and specificity are optimized significantly and consistently over different cross validation approaches. The implementation and evaluation has been done on a subset of the large UCI 130 hospital diabetes dataset. The performance measures of ensemble learners are compared to the base machine learning classification algorithms such as Naive Bayes, Logistic Regression, k Nearest Neighbor, Decision Trees and Support Vector Machines.
ContributorsBahl, Neeraj Dharampal (Author) / Bansal, Ajay (Thesis advisor) / Amresh, Ashish (Committee member) / Bansal, Srividya (Committee member) / Arizona State University (Publisher)
Created2017