Search Constraints

Start Over
Filtering by: Creator Kent, Adam J. R. Remove constraint Creator: Kent, Adam J. R. Committee Member Fern, Alan Remove constraint Committee Member: Fern, Alan Degree Name Master of Science (M.S.) Remove constraint Degree Name: Master of Science (M.S.)

Search Results

  • Methods for Detection and Recovery of Out-of-Distribution Examples

    Creator:
    Ratzlaff, Neale J.
    Abstract:
    Deep neural networks currently comprise the backbone of many applications where safety is a critical concern, for example: autonomous driving and medical diagnostics. Unfortunately these systems currently fail to detect out-of-distribution (OOD) inputs and can be prone to making dangerous errors when exposed to them. In addition, these same systems...
    Resource Type:
    Masters Thesis
    Full Text:
    estimation (Adam), and many others. These algorithms do not fundamentally deviate from GD, so they will not

  • Toward Disentangling the Activations of the Deep Networks via Low-dimensional Embedding and Non-negative Factorization

    Creator:
    Khorram, Saeed
    Abstract:
    In this thesis, we introduce a novel Explanation Neural Network (XNN) to explain the predictions made by a deep network. The XNN works by embedding a high-dimensional activation vector of a deep network layer non-linearly into a low-dimensional explanation space while retaining faithfulness i.e., the original deep learning predictions can...
    Resource Type:
    Masters Thesis
    Full Text:
    M ,S 1 2 ‖X −MS‖2F (2.8) subject to Mij ≥ 0, Sij ≥ 0 ∀i, j where r is the decomposition rank and

  • Counterfactual States for Atari Agents via Generative Deep Learning

    Creator:
    Olson, Matthew L.
    Abstract:
    Although deep reinforcement learning agents have produced impressive results in many domains, their decision making is difficult to explain to humans. To address this problem, past work has mainly focused on explaining why an action was chosen in a given state. A different type of explanation that is useful is...
    Resource Type:
    Masters Thesis
    Full Text:
    an Adam optimizer, with parameters α = 1−4, β1 = 0, β2 = .9. These networks were typically trained

  • Adaptive Movement Intent Decoding for Intuitive Control of Neuroprostheses

    Creator:
    George, Damon A.
    Abstract:
    Movement intent decoders, which interpret volitional movement intent from human bioelectric signals, can be incorporated into modern neuroprostheses to offer people living with limb loss or paralysis the potential to regain their lost motor control. Machine learning methods have become the research standard for continuous decoders with high degrees of...
    Resource Type:
    Masters Thesis
    Full Text:
    ) d,on(t1, t2) (3.9) L (k) d,on(t1, t2) = J(x̄d[t1, t2],�✓[n](z[t1, t2])d) r✓d[n]L (k) d,on(t1, t2

  • Agile Bipedal Maneuvers Optimized then Learned from Single Rigid-Body Model Dynamics

    Creator:
    Batke, Ryan
    Abstract:
    As bipedal robots move ever closer to being integrated into all manner of real world envi-ronments there is a necessity to push their dynamic capabilities to meet or exceed those of humans and animals. Advancements must be made to address ordinary challenges that arise everyday in the same environments that...
    Resource Type:
    Masters Thesis
    Full Text:
    (or control strategy) π that maximizes the expected return or cumulative discounted return R, defined

  • Protein secondary structure prediction using conditional random fields and profiles

    Creator:
    Shen, Rongkun
    Abstract:
    Protein secondary structure prediction plays a pivotal role in predicting protein folding in three-dimensions. Its task is to assign each residue one of the three secondary structure classes helix, strand, or random coil. This is an instance of the problem of sequential supervised learning in machine learning. This thesis describes...
    Resource Type:
    Masters Thesis
    Full Text:
    extensive help during this research. Thanks to Thomas G. Dietterich and Adam J. Ashenfelter for the core

  • Action Selection for Heterologous Control of a Continuous-Output Myoelectric Prosthesis

    Creator:
    Sylwester, Sean T.
    Abstract:
    Significance: Movement intent decoding algorithms can interpret human bioelectrical signals to control prosthetic limbs with many degrees of freedom (DOFs). This work involves decoding volitional movement intent from surface electromyogram (sEMG) signals to control prosthetic arms. To train these algorithms, patients flex their muscles to “follow” a movement prompt, and...
    Resource Type:
    Masters Thesis
    Full Text:
    samples: ei,j = 1 Nr Nr∑ r=1 √ 1 |Ni| ∑ n∈Ni ( xji,r[n]− x̄ j i,r[n] )2 e [0,1] i,j = ei,j

  • Efficient Incremental Panorama Reconstruction from Multiple Videos

    Creator:
    Feng, Zhongyuan
    Abstract:
    Constructing a panorama from a set of videos is a long-standing problem in computer vision. A panorama represents an enhanced still-image representation of an entire scene captured in a set of videos, where each video shows only a part of the scene. Importantly, a panorama shows only the scene background,...
    Resource Type:
    Capstone Project
    Full Text:
      (3.7) The cost function is defined such that C = ∑ ij ∑ k(r k ij) 2, where rkij = p k ij − f n j

  • Decentralized Multi-Robot Exploration with Sparse Communication through Deep Imitation Learning

    Creator:
    Senewiratne, Yomal K.
    Abstract:
    We present a method for decentralized, multi-robot exploration in adverse environments where communication is minimal. A key conceptual feature of our method is enabling implicit coordination between robots by training a Convolutional Neural Network (CNN) as a heuristic for planning using Monte Carlo Tree Search (MCTS). Our method consists of...
    Resource Type:
    Masters Thesis
    Full Text:
    location vt and and ground truth map Gj in trial j as input to produce training data for the CNN. The CNN

  • Open Set Learning with Counterfactual Images

    Creator:
    Neal, Lawrence W.
    Abstract:
    In open set recognition, a classifier must label instances of known classes while detecting instances of unknown classes not encountered during training. To detect unknown classes while still generalizing to new instances of existing classes, this thesis introduces a dataset augmentation technique called counterfactual image generation. This approach, based on...
    Resource Type:
    Masters Thesis
    Full Text:
    We treat Cp : R n → {0, 1} as an indicator function with the value 1 if p is true. Given the optimal