Encrypting and decrypting text files using a software-simulated Enigma Machine without the plugboard (A WWII German Encryption Machine). The encryption as well as decryption goes through a process of 7 stages simulating a series of 6 rotors and a reflector each with the 26 letters of the alphabet. A particular setting of rotor positions determines how encrypting or decrypting is carried out.

The System is designed following a use-case oriented architecture and implemented following an object-oriented approach (using Java) where different actors (roles) (passenger, conductor, cashier.) trigger different scenarios (actions and reactions). A typical scenario would be when a traveler (The passenger: use-case) requests a ticket, the system (The cashier: use-case) checks for availability with regards to time, departure location, destination location, and seat availability. If a ticket is available, a reservation is made. At the time of boarding, the traveler presents the ticket and the system (The conductor: use-case) verifies its validity. If the ticket is valid, the traveler boards the train to destination. When conditions are not satisfied, handlers to those events are also implemented. The system is backed with a MySQL database for CRUD operations.

An Android application that allows a user to upload or take a picture of a human face. After this initial phase, the file is sent to a Face & Emotions detection API (Used the Microsoft Azure Platform). Once processing is complete by the API, results will be sent back to the user for display (in JSON). Among the results are: (1) clear detection of the face (if it belongs to a human and it is visible) by placing a colored rectangle around it, (2) Emotions detected from the face such as happiness, anger, fear, neutrality, ...etc.

This project solves the Minimum Spanning Tree (MST) problem using Kruskal’s Algorithm with path compression. To achieve our goal, data for weighted graphs with (n) vertices and (m) edges are given in text files. An MST, if it exists, is a subgraph containing (n – 1) edges such that when their weights are added together, they give the minimum possible cost needed to visit all vertices exactly once without creating any cycles. An MST is not possible to compute if the graph is not connected as the algorithm runs out of edges to consider before satisfying the above condition. The implementation of Kruskal’s algorithm uses a disjoint-set data structure implemented using an array of integers called “sets”. In this data structure at start, the cell at index “i” is the single representative (or label) of the element “i”. Thus, we make (n) disjoint subsets where one of them contains one single vertex. As the subsets merge together, if the index “i” is not the representative of its subset, we go through the subsets until we find its representative. Once this operation is achieved, we proceed to path. The compression of the path that led to our target can be achieved by introducing a caching mechanism into the find operation where all elements in the path to the representative will, after that, point to the target label. This way it speeds up the subsequent find operations and speeds up the overall running time of Kruskal’s algorithm. At the end, if an MST is computed, all edges that make it will be displayed along with their total weight.