An academic paper on an economic parameterization for positive parahermitian matrix functions to obtain a state-space solution for μ-analysis, via reformulation as a convex optimization problem. It avoids the use of a frequency-dependent curve fitting approach, by Laguerre parameterization and spectral factorization within a LMI framework. The algorithm gives a frequency-independent μ upper-bound with less number of decision variables, a path to obtain μ without Matlab's prioprietary mu, and an advantage of less computational time. Two Matlab implementations demonstrate the algorithm's effectiveness. [Matlab, Robust Control, Algorithm]
A microservice implementation that exposes a Maths operator: Addition [for single digits], Average/Maximum/Minimum/Summation/Sorted computes [for numeric arrays], via REST API endpoints and deployed to a Kubernetes cluster. [Microservices, Go, gRPC, Protobuf, Docker, Kubernetes]
A classification algorithm that applies scikit-learn to the problem of optical character recognition of hand-written numerical digits. The code focuses on identifying the numerical character in the image, without attempting to locate the image itself. Test/validation data comes via scikit-learn's preformatted digits. The approach uses a test harness to initially spot-check different classification models: LR, LDA, KNN, NB, SVM, CART, before making a final prediction based on the KNN model. [K-Nearest Neighbors, Seaborn, Python, Numpy, Matplotlib, Pandas, Scikit-Learn]
A simple Golang implementation of a text-based adventure game: * in which a player is able to choose alternative paths * that renders to a webserver base UI * uses flags (`-port`, `-file`, `-switch`, `-template`) for user defined attributes * while implementing the `functional options` design pattern [Go, Command-Line Flags, JSON, Url Parser, HTML, CSS, Design Pattern]
A simple Scikit-Learn implementation of Iris Flower classification i.e. supervised learning, that uses the `iris flowers` within the `iris dataset`. The approach uses a test harness to initially spot-check different classification models: LR, LDA, KNN, NB, SVM, CART, before making a final prediction based on the KNN model. [Iris Dataset, K-Nearest Neighbors, Seaborn, Python, Numpy, Matplotlib, Pandas, Scikit-Learn]
A golang implementation of GraphQL server, deployed to Google App Engine serverless platform. The GraphQL server is hosted at https://graphqlserver-259904.appspot.com/graphql. The server allows for mutations (to create users/posts) and query (of users/posts) via either a Post or Get method. [GraphQL, Go, Google App Engine]
An algorithm that implements gradient descent and backpropagation training methods, from first principles - without recourse to frameworks. The code performs training and accuracy prediction using the MNIST dataset. [Python, Numpy, Matplotlib, Scikit-Learn]
A simple Golang implementation of a Url Shortener that: * examines path of incoming request * determines if re-direction is required * uses flags (`-yaml`, `-json`, `-sql`) to source the required content from files or database instead of inline strings [Go, PostgreSQL, Command-Line Flags, YAML, JSON, Url Parser, ]
A monitoring application that displays a Youtube channel's statistic [Views count, Subscribers count, Videos count]. It consists of a simple frontend that is feed with JSON data by a backend through a websocket connection. The implementation leverages Youtube's API, Docker image [generated via Dockerfile] and endpoints implemented using Gorilla Mux. [Go, API, Docker, Websockets, JS, JSON, Gorilla Mux]
A simple Golang implementation of a Quiz that: * asks a user simple maths questions * validates their answer * computes their % correct answer * times the user in between questions * sources both the question and answers from a CSV file [Go, GoRoutines/Channels, Command-Line Flags, CountDown Timer, CSV]
A gRPC based web service that consumes Gravatar's API to generates URLs mapped to a globally unique email address and personalized avatar. Test coverage only covers the `hasher` package. [Go, gRPC, Protocol Buffers, Gravatar]
A simple python implementation of human facial recognition, based on Python Imaging Library (PIL) that uses pre-prepared test/stock images at runtime. [Python, PIL (Python Imaging Library)]
A simple scikit-learn implementation of a Gaussian Mixture Model i.e. a primer about unsupervised learning: clustering, based on the Iris dataset. [Python, Numpy, Matplotlib, Seaborn, Scikit-Learn]
A simple scikit-learn implementation of Principal Component Analysis i.e. a primer about unsupervised learning: dimensionality reduction, based on the Iris dataset. [Python, Numpy, Matplotlib, Seaborn, Scikit-Learn]
A simple scikit-learn implementation of Naive Bayes i.e. a primer about supervised learning: classification, based on the Iris dataset. [Python, Numpy, Matplotlib, Seaborn, Scikit-Learn]
A simple scikit-learn implementation of Linear Regression i.e. a primer about supervised learning: regression analysis, using randomly generated training and testing data. [Python, Numpy, Matplotlib, Seaborn, Scikit-Learn]
A simple Chat application based on a React (frontend) and Go (backend) that communicate via websockets. [React, Websocket, Go REST API, Go Modules, Go Routines, Gorilla Mux, Node Sass, Dockerfile]
A bare-bones Blog with authentication [login, logout, register] features and tests coverage. [Flask, Pytest, SQLite, HTML, CSS, Jinja2]
