Loading...

Pruning Machine Learning Models by Sparse Representation

Khorashadizadeh, Amir Ehsan | 2020

512 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 52812 (05)
  4. University: Sharif University of Technology
  5. Department: Electrical Engineering
  6. Advisor(s): Babaiezadeh, Massoud
  7. Abstract:
  8. In recent years, Machine Learning models have been developed in Signal Processing, Computer Vision and Neuroscience areas. There are two categories of Machine Learning models which are supervised and unsupervised learning models. Regression and classification problems are two popular problems examples of supervised learning models. From unsupervised learning problems, we can mention the clustering problem. Support Vector Regression (SVR), Decision Tree Regression and Bagging Ensemble Regression models are some important models of the regression problem. For classification problems, we can also mention to Support Vector Classification, Decision Tree Classification, and Bagging Ensemble Classification models. In this thesis, we plan to focus on classification and regression problems.Designing a Machine Learning model consists of two stages that are training and test stages. In the training stage, the Machine Learning model should be trained on the training dataset. then, the performance of the model is to be evaluated by the test dataset. The performance of the model can be evaluated by some factors such as the processing speed, the required memory and the performance of the model on the test dataset. In this thesis, we plan to enhance the performance of the Machine Learning model. Enhancing the model performance means increasing the processing speed, reducing the required memory and improving the model performance on the test dataset.Currently, Sparse Representation is one of the popular subjects in Signal Processing. Sparse Signal Representation aims to represent a signal based on a linear combination of a few base signals. Each of the base signals is called an atom and the combination of them is called a dictionary.Pruning is a technique to improve a Machine Learning model performance. Pruning a Machine Learning model means to remove as much as parameters of the model such that the performance of the model will not be affected. In this thesis, two novel algorithms for pruning the Bagging Ensemble Regression and Classification models by Sparse Representation have been presented. As we will see in this thesis, not only pruning reduces the size of ensemble models, but also, it improves the model performance over test dataset
  9. Keywords:
  10. Machine Learning ; Sparse Representation ; Pruning Method ; Bagging Ensemble Regression Analysis ; Pruning Machine Learning Models ; Bagging Ensemble Classification

 Digital Object List

 Bookmark

No TOC