This repository contains a C++ program that implements a neural network process. The program reads weights from a configuration file, performs forward propagation using those weights, and computes the output based on the provided weights and data. The program utilizes multi-threading and inter-process communication using pipes.

Make sure you have the following installed on your system:

• C++ compiler (supporting C++11)
• POSIX threads (pthread) library

1. Clone the repository to your local machine:

git clone <https://github.com/Vaneeza-7/Neural-Network-implementation-using-OS-Concepts.git>

2. Compile the source code using the C++ compiler:

g++ new.cpp -o neural_network -lpthread

3. Run the program:

./neural_network

Just give the path to the configuration file and the prgram reads the file for weights and input. For a different input, just change the configuration file and the code will run accordingly. The program will create multiple processes and threads to simulate the neural network. Each process represents a layer, and each thread represents a neuron. The calculations for each neuron are performed concurrently using threads. The output of each neuron is printed to the console.

• The program uses a pthread_mutex_t mutex to synchronize access to shared data.
• It reads the weights from the Configuration.txt file and stores them in appropriate data structures.
• It also reads the input data from the Configuration.txt file.
• Performs forward propagation using the provided weights and input data.
• Computes the output of the neural network.
• The n1 function represents a neuron in the input layer. It performs calculations based on the weights and input values.
• The n2 function represents a neuron in the hidden layers. It performs calculations using the output of input layer and its weights.
• The n3 function represents a neuron in the outer layer. It gives the output using the output of hidden layers and reads the final result from a pipe.
• Processes are created using fork(), and threads are created using pthread_create().
• Pipes are used for inter-process communication. Each process/thread reads data from the previous process/thread and writes its output to the next process/thread.
• The program prints the process IDs to the console for reference.

• Vaneeza
• Rabail

This project is licensed under the MIT License.