DSCI 599 - Optimization Techniques for Data Science - Fall 2023 - USC

This repository contains solutions to assignments, including programming tasks, project updates, and final project reports. Each assignment's solution is organized in separate folders, and clear instructions are provided for submission.

Tip

Before exploring the materials, take a moment to review the license and disclaimer for responsible utilization. The repository covers various topics, providing valuable insights into Optimization Techiniques and solution to assignments.

Course Details:

• Course Name: DSCI 599 - Optimization Techniques for Data Science
• Instructor: Prof. Satish Thittamaranahalli Ka
• Semester: Fall 2023

About

A 4-unit course designed to explore various optimization techniques drawn from algorithms, artificial intelligence, machine learning, and mathematics. The course aims to provide students with a comprehensive understanding of optimization algorithms and their applications in solving problems relevant to data science. Topics covered include linear programming, convex programming, machine learning paradigms (supervised, unsupervised, reinforcement learning), graph theory, heuristic search, constraint programming, and more.

Caution

Please note that this repository serves as a reference guide and should be utilized as a tool for learning and comprehension. It's paramount to refrain from engaging in any activities associated with plagiarism. Embrace the wealth of knowledge herein to enhance your understanding and augment your skill set in the field of machine learning algorithms.

Table of contents

Assignment	Topic Covered
Homework 1	Dynamic Programming, Local Search, K-Means Clustering, Expectation Maximization, and FastMAP
Homework 2	Guassian Mixture Models (GMMs), Multivariate Guassian, Principal Component Analysis (PCA), Dimensionality Reduction, Perceptron Learning Algorithm and Constraint Satisfaction Problems (CSPs)
Homework 3	k-consistency, Backtracking search, Look-Ahead, and Dynamic Variable/Value Ordering in CSPs, Simple Temporal Problems (STPs), and Bayesian Networks
Homework 4	Dynamic Belief Networks (DBNs), Markov Models (MMs), Hidden Markov Models (HMMs), Weighted Constraint Satisfaction Problems (WCSPs), Graph Theory, Feed-Forward Neural Networks and Polynomial Identity Checking
---	---
Final Project	Exploring Adversrial learning strategies for Sequence Labeling Tasks

Authors

1. Kayvan Shah | MS in Applied Data Science | USC

LICENSE

This repository is licensed under the BSD 5-Clause License. See the LICENSE file for details.

Disclaimer

_{The content and code provided in this repository are for educational and demonstrative purposes only. The project may contain experimental features, and the code might not be optimized for production environments. The authors and contributors are not liable for any misuse, damages, or risks associated with the use of this code. Users are advised to review, test, and modify the code to suit their specific use cases and requirements. By using any part of this project, you agree to these terms.}