University Computer Science Projects Portfolio

Welcome to this repository showcasing a diverse collection of university projects. Each project targets a different paradigm or aspect of computer science, ranging from functional programming and machine learning to network security and interactive applications.

Table of Contents

1. Functional & Logical Programming (FLP) Assignment
2. Multilayer Perceptron (MLP) Regressor
3. Stock Price Dataset Analysis
4. Quiz Game
5. Network Scanner with Scapy

---

1. Functional & Logical Programming (FLP) Assignment

This project simulates the creation, processing, and analysis of a system log file, demonstrating how the same data problem can be approached from both Functional and Logical paradigms.

Core Workflow

• Log File Generation: Automatically generates a fake log file containing 1,000 entries. Each entry includes a timestamp, a log level (INFO, ERROR, WARNING, or DEBUG), and a context-specific message.

• Data Extraction & Processing: Parses the generated log file to calculate:

• The total number of log entries per hour (24-hour breakdown).

• The total number of entries for each log level.

• The average length of log messages.

• Data Visualization: Plots the extracted statistics using bar charts (activity by hour), a pie chart (distribution of log levels), and a histogram (distribution of message lengths).

Paradigm Perspectives

🔹 Functional Programming Perspective

The program processes data through a stateless workflow, emphasizing:

• Pure Functions: Tasks like generating messages and parsing lines have predictable outputs and zero side effects.
• Data Transformation: Data flows through mappings and transformations rather than altering structures in-place.
• Immutability: Log entries and counters are created fresh at each stage instead of mutating a global state.
• Composability: Reusable functions are chained together from file I/O to final visualization.

🔹 Logical Programming Perspective

Alternatively, the log analyzer can be conceptualized through formal logic:

• Collection of Facts: Every log line is treated as a structured fact within a database.
• Rule-Based Reasoning: Logic rules can be defined to infer system anomalies (e.g., flagging sudden spikes in ERROR levels).
• Declarative Intent: Focuses on what conditions to find (e.g., "Find all hours where warnings exceed 50") rather than step-by-step math implementation.

---

2. Multilayer Perceptron (MLP) Regressor

This project demonstrates how a Neural Network can approximate complex, non-linear functions using the MLPRegressor class from scikit-learn.

Implementation Details

• Data Generation: Creates synthetic training data based on a noisy sine wave ($y = \sin(X) + \text{noise}$) where $X$ ranges from 0 to 5.
• Model Architecture: A Multilayer Perceptron consisting of two hidden layers (100 neurons in the first layer, 50 neurons in the second layer) configured to train for a maximum of 1,000 iterations.
• Training & Prediction: The model iteratively adjusts its internal weights to minimize the error between predictions and the true target values. Once trained, it predicts values across a smooth continuum ($x_{\text{new}}$ from 0.0 to 5.0).
• Visualization: Generates a plot overlaying the original noisy training points against the smooth regression line learned by the model to evaluate its accuracy.

---

3. Stock Price Dataset Analysis

Building upon the concepts explored in the MLP regression model, this project applies neural network-driven regression to a real-world scenario: analyzing and forecasting trends within a sample stock price dataset.

---

4. Quiz Game

🎮 Just for fun! An interactive, text-based terminal quiz game designed as a lighthearted application of basic programming constructs, conditional logic, and user input handling.

---

5. Network Scanner with Scapy

A surface-level networking script that utilizes the Scapy library to discover active devices on a local network and pull hardware vendor insights.

Architecture & Mechanics

1. Vendor Resolution (get_vendor)

• Takes a MAC address as an input and targets the external MAC Vendors API (https://api.macvendors.com/).
• Issues an HTTP GET request to fetch the hardware vendor. Returns "Unknown" if the API is unreachable or the MAC address is unrecognized.

2. Network Scanning (scan_network)

• Constructs and broadcasts ARP (Address Resolution Protocol) request packets across a specified IP range in CIDR notation (Default: 192.168.100.1/24).
• Uses Scapy's srp() function to send packets and listen for responses with a 2-second timeout window.
• Active devices respond with their MAC addresses, which are paired with the IP address and matched to a vendor name.

Sample Output Format

The results are displayed in a clean, tabular structure directly within the console:

IP Address	MAC Address	Vendor Name
192.168.100.1	00:1A:2B:3C:4D:5E	Cisco Systems
192.168.100.15	AA:BB:CC:11:22:33	Apple Inc.
192.168.100.102	4E:55:63:A1:B2:C3	Unknown