Randomized Optimization (ML Assignment 2)
نویسنده
چکیده
I implemented Randomized Hill Climbing (HC), Simulated Annealing (SA), and Genetic Algorithms (GA) in Tensorflow/DEAP and used each approach to train a very simple neural network on my dataset. The neural network is fully-connected and has a single hidden layer with 100 units that use tanh activations. The loss function is defined as the mean categorical cross entropy over the 10 labels. Since this architecture differs somewhat from the one used in Project I (which also used a powerful optimizer), I also ran a (backpropagation) baseline using gradient descent.
منابع مشابه
A Greedy Randomized Adaptive Search Procedure for the Quadratic Assignment Problem
A greedy randomized adaptive search procedure (GRASP) is a randomized heuristic that has been shown to quickly produce good quality solutions for a wide variety of combinatorial optimization problems. In this paper, we describe a GRASP for the quadratic assignment problem. We review basic concepts of GRASP: construction and local search algorithms. The implementation of GRASP for the quadratic ...
متن کاملA Framework for Randomized Optimization Algorithms and Its Application to the Wavelength Assignment Problem
We propose a library-framework for optimization algorithms which perform a step-wise, randomized search in the solution space. The optimization problem has to be formulated once, then any of these algorithms can work on this problem. We show that the NP-complete problem of assigning wavelengths to paths in optical networks can be solved (sub-optimally) by framework-based algorithms, and that th...
متن کاملScalable, Balanced Model-based Clustering
This paper presents a general framework for adapting any generative (model-based) clustering algorithm to provide balanced solutions, i.e., clusters of comparable sizes. Partitional, model-based clustering algorithms are viewed as an iterative two-step optimization process—iterative model re-estimation and sample re-assignment. Instead of a maximum-likelihood (ML) assignment, a balanceconstrain...
متن کاملRandomized Parallel Algorithms for the Multidimensional Assignment Problem
The Multidimensional Assignment Problem (MAP) is a combinatorial optimization problem arising in diverse applications such as computer vision and motion tracking. In the MAP, the objective is to match tuples of objects with minimum total cost. Randomized parallel algorithms are proposed to solve MAPs appearing in Multi-Sensor Multi-Target applications. A parallel construction heuristic is descr...
متن کاملParallel Strategies for Grasp with Path-relinking
A Greedy Randomized Adaptive Search Procedure (GRASP) is a metaheuristic for combinatorial optimization. It usually consists of a construction procedure based on a greedy randomized algorithm and local search. Path-relinking is an intensification strategy that explores trajectories that connect high quality solutions. We analyze two parallel strategies for GRASP with path-relinking and propose ...
متن کامل