Multiple faces are classified on the basis of probabilistic density comparison of different faces. Deep Learning Techniques, such as Deep Convolution Networks and Matrix Augmentation are used.

A website depicting a real-life scenario of SBI(Bank in India) while distributing the computing power to multiple nodes in the environment. Motivation behind this project was to create multi-noded system(Distributed Environment).

Various components of the operating systems were modified according to college course requirements.

A basic course project where the several galaxies were classified into spiral or normal based on their radius and emission spectrum. K-Means Clustering was the base algorithm for clustering the galaxies into different types.

This was a phone number and information storage with retrieval project. All basic and complex actions, such as input, delete, modify were implemented. Also levels of leverages were defined, they were User-Level and Admin-Level.

Authentication, Authorization with jwt token methodology.

A customizable authentication middleware for user data security. Email based authentication method are predefined.

A multi nod-ed environment with distributed computing operations. Distributed crawler with politeness feature for fetching documents from web. Efficient word indexing for user-query processing with recent search caching. This search engine also has many api based calls for section discretization.

People Interaction through multi-level commenting engine. Deep comments are processed with custom foreign key written in Python and Stored in Sqlite3. Depth of Comment can be changed with slight modification in source code for custom choices.

A general portal where questions(queries) can posted, with a stable answering environment. The above (1) and (3) are(can) be embedded in this environment.

Machine learning based back-end potential theft monitoring core. Main algorithms used here is Naive based classification method to detect anomaly based on feature value irregularity. A clear theft is clearly detected in scenarios of high traffic or some bot based attack.

Reverse proxy built on Nginx, Apache. Shell used for request logging. Python is used for detecting anomaly using Machine Learning algorithms on the basis of features such as IP address.

Inherited from above (6) with some additional utilities. This Load-Balancer has option for normal request routing and persistence(mainly for data processing requests) request routing. Various load balancing paradigms such as round robin, cpu power, admin choice, random, etc are possible. 429(Request Overloading) Error can also be handled by temporary(or permanent, based on overloading factor) blocking the user access.

An VPN for request hopping to provide user privacy. Shell used for logging user Client Requests. Machine Learning algorithms used for detecting anomalies in client fetch event.