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 53 for “"Frank Wolfe"”.
-
Momentum for the Frank Wolfe Method
… sparse) solutions. Conditional gradient, aka Frank Wolfe (FW) algorithms, have well-documented merits in machine learning and signal processing applications that involves minimizing a loss function with constraints. Compared to projection based methods, one of the key benefits is that FW …
-
Frank-Wolfe Optimization for Dominant Set Clustering
… we study the StQP from the perspective of the Frank-Wolfe (FW) algorithm, and variants thereof, and relate them to replicator dynamics – a method commonly used for computing dominant sets. We consider standard FW, pairwise FW, and away-steps FW, and conclude that all variants perform similarly …
-
Frank-Wolfe Methods for Optimization and Machine Learning
In Chapter 2, we present the Frank-Wolfe algorithm (FW) and all necessary background material. We explain the projection-free and sparsity properties of the algorithm, provide motivation for real-world problems, and analyze the convergence rates and a lower bound on the complexity. In Chapter 3, we …
-
Algorithm robustness and its applications : Frank-Wolfe and factorized gradient
… and second part, we study the robustness of Frank-Wolfe type methods in two important settings: robustness against the perturbation created by asynchronous updates, and then robustness to the corruption produced by the injection of adversarially corrupted data in the training set. In the …
-
Solving Saddle Point Formulations of Linear Programs with Frank-Wolfe
… formulations of linear programs, named FWLP (Frank-Wolfe Linear Programming). We provide some theoretical results regarding the behavior of our algorithm, however no convergence guarantees are provided. Numerical investigations suggest that our algorithm has error O(1/sqrt(k)) after k …
-
Parallel Implementations of the Frank-Wolfe Algorithms for the Traffic Assignment Problem
… how traffic will flow within the network. The Frank-Wolfe algorithm is a traditional solution method for this optimization problem, but it has been characterized by its slow rate of convergence and poor computational performance. This thesis examines and implements several modern advancements …
-
Understanding Modern Techniques in Optimization: Frank-Wolfe, Nesterov's Momentum, and Polyak's Momentum
… algorithms, the popular ones include the Frank-Wolfe method, Nesterov's accelerated methods, and Polyak's momentum. While theoretical analysis of the Frank-Wolfe method and Nesterov's methods are available in the literature, the analysis can be quite complicated or less intuitive. Polyak's …
-
JAVA Animated Software for Teaching the Frank-Wolfe Algorithm for Static Traffic Network Equilibrium
The popular Frank-Wolfe (FW) algorithm for solving the network equilibrium problems plays an important role in transportation simulation. Not only has the basic Frank Wolfe algorithm been studied, but also other variations of the FW algorithm (such as Conjugate Frank Wolfe and Bi-Conjugate Frank …
-
On Variants of Sliding and Frank-Wolfe Type Methods and Their Applications in Video Co-localization
… is to explore some variants of sliding and Frank-Wolfe (FW) type algorithms, analyze their convergence complexity, and examine their performance in numerical experiments. We achieve three accomplishments in our research results throughout this dissertation. First, we incorporate a linesearch …
-
Optimization Methods for Semi-Supervised Learning
… the trace norm constrained problem utilizing the Frank-Wolfe algorithm, which only requires calculating the leading singular vector pair, providing an order of magnitude improvement on the iteration complexity. However, the Frank-Wolfe algorithm empirically has very slow convergence and in …
-
Methods for barcode analysis in optical DNA mapping
… Models.Paper V proposes the use of Sliding Frank-Wolfe methods for sparse-labeled single-frame ODM.Paper VI extends the Sliding Frank-Wolfe methods from the analysis of single frame barcodes to multi-frame setting, where barcodes over multiple time-frames are averaged to improve the …
-
A reduction framework for approximate extended formulations and a faster algorithm for convex optimization
… gradient type algorithms (also known as Frank-Wolfe type methods) are very important in theory as well as in practice due to their simplicity and fast convergence. We show how we can eliminate the linear optimization step performed in every iteration of Frank-Wolfe type methods and …
-
Faster Conditional Gradient Algorithms for Machine Learning
In this thesis, we focus on Frank-Wolfe (a.k.a. Conditional Gradient) algorithms, a family of iterative algorithms for convex optimization, that work under the assumption that projections onto the feasible region are prohibitive, but linear optimization problems can be efficiently solved over the …
-
Selecting and scheduling of improvements in urban transportation networks using metaheuristics
… the costs of network flows determined with the Frank-Wolfe algorithm, which is modified to consider intersection flows and delays. Intersections are modelled with pseudo-links. The methods used for choosing the optimal schedule of project improvements are: Ant Colony Optimization, Simulated …
-
Gauges and Optimization over Structured Sets
… this, a negative result on the inability of Frank-Wolfe style methods to benefit from smoothness is presented.
-
Work Zone Analysis Under Detour Settings
… the traffic. To solve for User Equilibrium, the Frank-Wolfe algorithm is implemented via Matlab. The total delay cost is then converted to a monetary value using value of time (VOT). The work zone set-up cost and accident cost relate to number of work zone segments and total length of work zone. …
-
Integrating Congestion Pricing And Transit Investment Planning
… is an integration of a genetic algorithm and the Frank-Wolfe method. The method and solution procedure is applied to an illustrative example.
-
Bipartite Matching and Routing with Congestion Costs: A convex approach to robot task assignment and the multi-agent pathfinding problem
… reach Wardrop equilibria. Finally, we present a Frank-Wolfe-based algorithm combined with shortest path algorithms such as Dijkstra's and A* algorithms and matching algorithms such as the Hungarian (Kuhn-Munkres) algorithm to find the optimal solution while considering congestion in the system. …
-
Optimal stochastic and distributed algorithms for machine learning
… simplex constraint can be handled by the classic Frank-Wolfe method. The proposed stochastic Frank-Wolfe methods achieve comparable or even better accuracies than state-of-the-art batch and online kernel SVM solvers, and are significantly faster. The last part investigates the problem of …
-
Prediction in High-Dimensional Statistical Modelling Using Greedy Algorithms and Information Theoretic Criteria
… Pursuit (OMP), Relaxed Matching Pursuit (RMP), Frank-Wolfe Algorithm (FWA) and Constrained Matching Pursuit (CMP). We show how twelve information theoretic (IT) criteria can be used jointly with the greedy algorithms. As part of this e ort, we derive new theoretical results that allow us to …
Page 1 of 3