🚀 SPPU-BE-COMPUTER-LAB-CODE-SEM-8 🚀

Welcome to the repository for SPPU BE Computer Semester 8 Lab Code! This repository contains Python-based implementations of lab exercises for two subjects: High Performance Computing (HPC) , Deep Learning (DL) , Software Defined Networks (SDN) , Business Intelligence (BI) . Below, you will find a detailed description of the topics covered in each subject along with small descriptions.

---

📋 Table of Contents

1. High Performance Computing (HPC)

   • 1. Parallel BFS and DFS using Python Multiprocessing
   • 2. Parallel Bubble Sort and Merge Sort using Python Multiprocessing
   • 3. Parallel Reduction for Min, Max, Sum, and Average
   • 4. CUDA Programs

2. Deep Learning (DL)

   • 1. Linear Regression using Deep Neural Networks
   • 2. Binary Classification using Deep Neural Networks
   • 3. Convolutional Neural Network (CNN)
   • 4. Recurrent Neural Network (RNN)

3. Software Defined Networks (SDN)

   • 1. Mininet Network Emulation Environment Setup
   • 2. OpenFlow Controllers: POX and Floodlight
   • 3. SDN Firewall Using Mininet and POX
   • 4. Mininet POX Router
   • 5. Emulating and Managing a Data Center via a Cloud Network Controller

4. Business Intelligence (BI)

   • 1. Legacy Data Import and Load to Target System
   • 2. ETL Process for Constructing a Database in SQL Server
   • 3. Data Cube Implementation: ROLAP, MOLAP, and HOLAP
   • 4. Data Warehouse Analysis with Microsoft Excel
   • 5. Data Classification and Clustering Assignment

---

🖥️ High Performance Computing (HPC)

1. Parallel BFS and DFS using Python Multiprocessing

• Objective: Design and implement parallel versions of Breadth First Search (BFS) and Depth First Search (DFS) using Python's multiprocessing.
• Description:
  • BFS and DFS are implemented on a tree or an undirected graph.
  • Parallelization is achieved using Python's multiprocessing module to distribute tasks across multiple processes.

OUTPUT

2. Parallel Bubble Sort and Merge Sort using Python Multiprocessing

• Objective: Implement parallel versions of Bubble Sort and Merge Sort using Python's multiprocessing.
• Description:
  • Compare the performance of sequential and parallel versions of the algorithms.
  • Measure execution time to analyze the speedup achieved through parallelization.

OUTPUT

3. Parallel Reduction for Min, Max, Sum, and Average

• Objective: Perform parallel reduction operations for computing Min, Max, Sum, and Average.
• Description:
  • Use Python's multiprocessing to perform parallel reduction on arrays.
  • Demonstrate how reduction operations can be efficiently parallelized.

OUTPUT

4. CUDA Programs

4.1 Addition of Two Large Vectors

• Objective: Implement vector addition for large vectors using CUDA Python (numba.cuda).
• Description:
  • Utilize GPU parallelism to add two large vectors.
  • Compare the performance with CPU-based implementation.

4.2 Matrix Multiplication using CUDA Python

• Objective: Implement matrix multiplication using CUDA Python (numba.cuda).
• Description:
  • Leverage GPU threads to perform matrix multiplication in parallel.
  • Analyze the performance improvement over sequential matrix multiplication.

OUTPUT

---

🧠 Deep Learning (DL)

1. Linear Regression using Deep Neural Networks

• Objective: Predict Boston housing prices using a Deep Neural Network (DNN).
• Description:
  • Use the Boston Housing Price Prediction dataset.
  • Build a DNN model to perform linear regression and predict house prices.

2. Binary Classification using Deep Neural Networks

• Objective: Classify movie reviews as "positive" or "negative" using a DNN.
• Description:
  • Use the IMDB dataset containing movie reviews.
  • Train a binary classification model to classify reviews based on text content.

3. Convolutional Neural Network (CNN)

3.1 Plant Disease Detection System using CNN

• Objective: Design a plant disease detection system using a CNN.
• Description:
  • Use a dataset of plant images with various diseases.
  • Train a CNN model to detect and classify plant diseases.

3.2 Fashion Clothing Classifier using MNIST Fashion Dataset

• Objective: Create a classifier to categorize fashion clothing using CNN.
• Description:
  • Use the MNIST Fashion dataset.
  • Train a CNN model to classify clothing items into categories like T-shirt, Dress, etc.

4. Recurrent Neural Network (RNN)

• Objective: Design a time series analysis and prediction system for Google stock prices using RNN.
• Description:
  • Use the Google stock prices dataset.
  • Train an RNN model to analyze historical data and predict future stock prices.

---

🛠️ Getting Started

Prerequisites

• Python Environment: For all programs, set up Python with libraries like TensorFlow, Keras, NumPy, Pandas, and Matplotlib.
• CUDA Toolkit: For CUDA programs, install the CUDA toolkit and compatible GPU drivers.
• Datasets: Download datasets from official sources (e.g., Boston Housing, IMDB, MNIST Fashion, Google Stock Prices).

Installation

1. Clone this repository:

   git clone https://github.com/yourusername/SPPU-BE-COMPUTER-LAB-CODE-SEM-8.git
   cd SPPU-BE-COMPUTER-LAB-CODE-SEM-8

2. Install dependencies:

   • For Python:

     pip install tensorflow keras numpy pandas matplotlib scikit-learn numba cudatoolkit

3. Run the programs:

   • Run Python scripts directly:

     python script_name.py

   • Or use Jupyter notebooks:

     jupyter notebook

---

Happy Coding! 🚀