Mahsatajik/NNDL

University of Tehran-Neural Networks and Deep Learning Spring 2020

Overview

This repository contains some of my projects from the Neural Networks and Deep Learning course I took during my Master's program at the University of Tehran. The course provided an in-depth understanding of the architecture of deep neural networks and the algorithms developed to extract high-level feature representations from data.

Course Content

Theoretical Foundations

• Backpropagation: Learning algorithm for training neural networks.
• Stochastic Gradient Descent (SGD): Optimization algorithm used for minimizing the cost function.

Practical Implementations

We gained hands-on experience building deep neural network models using Python and popular machine learning libraries, such as TensorFlow and Keras. Below are the types of neural networks we explored:

• Convolutional Neural Network (CNN): Used primarily for image recognition and classification tasks.
• Recurrent Neural Network (RNN): Suitable for sequential data and time-series analysis.
• Self-Organizing Maps (SOM): Unsupervised learning technique for clustering and visualization of high-dimensional data.
• Long Short-Term Memory (LSTM): A type of RNN effective for learning long-term dependencies.

Repository Structure

• README.md
• CA1_associative_memory
  • notebook.ipynb
  • description
  • report
• CA2_
  • notebook.ipynb
  • description
  • report
• CA3_RNNs
  • notebook.ipynb
  • description
  • report
  • pollution_dataSet.csv
  • pollution_dataSet.npy
  • pollution_dataSet1.csv

Acknowledgements

I would like to thank my professor and peers at the University of Tehran for their guidance and support throughout this course.