kasiotis Goto Github PK

bird-image-classifier-task

The objective of this project was to train a network that can accurately classify species of birds by using transfer learning (using weights pre-trained on ImageNet).

data-management-and-exploratory-data-analysis

This was a project that was part of my Data Science with AI Master's degree. As part of this project I had to follow the CRISP-DM methodology and use the ProjectTemplate library to carry out my EDA on a never seen before collection of datasets. The process included working to set the business objectives, understand the provided data, preparing it for analysis and carrying out the final statistical analysis. Furthermore, a 20-page report along with a presentation were produced to explain the details of the whole process and showcase findings.

data-structures-and-swing-components-project

This was a Java project that was part of my Undergraduate degree. As part of this project we were tasked with implementing a staff telephone directory that supported a range of operations. These operations had to be implemented with a variety of data structures and then get tested to find the ones that worked the best for each operation. Additionally, the directory also had to have a Graphical User Interface that had to be developed with the use of Swing Components.

do-to-done-flutter-app

language-model-task

Models capable of generating text are prevalent and present in many aspects of modern life. The objective of this project was to train a word-level language model using the text from the book “Poirot Investigates” by Agatha Christie.

oop-sandwich-builder

This was an Undergraduate project that I worked on where I implemented an Object-Oriented program in Java that would act as a sandwich builder. Multiple object were created to represent the various items that make up a sandwich. Prices were also included for each item and in the end when you built your custom sandwich, a receipt was printed with the prices for each item used and a total amount.

skin-cancer-mnist-ham10000

For this project, I set out to implement three different image classifiers and fit all of them with greyscale and RGB images to observe their differences in performance and find the one that works best. Additionally, I wanted to investigate the effect of the image colour spectrum on a range of classifiers. The classifiers that I set out to implement are the cnn (convolutional neural network), the lstm (Long short-term memory RNN), and the svm (Support vector machine).

sorting-algorithms

This was an Undergraduate project that I worked on where I was tasked with implementing and comparing Sorting Algorithms

terascope-supercomputer-performance-eda

A Terapixel visualization requires the use of systems that are far larger and more complex than the ordinary computer system. The goal of this investigation is primarily to delve into the matters of software system and hardware performance and operation of supercomputers that run Terapixel visualization applications.

test-driven-development

This was an Undergraduate project that I worked on. As part of this project we work with Java and we used a framework and JUnits to implemented a set of test cases that were later used to test our code implementation for a number of functions.

testing-and-debugging

This was an undergraduate project that required me to test some Java classes as defined by a software specification. Specifically I had look for faults in the code and other non-functional aspects such as examples of what I would consider bad software engineering practice.

vdm-modelling

This was an Undergraduate project where I worked with VDM-SL in Overture to complete a software modelling task.