Bakalis Dimitrios's repositories
Image-Classification-on-Brain-MRI
In this project i used a data set with images from brain MRI, to make a convolutional network (using Keras) to effectivly detect images that contain a tumor.
Analysis-of-Pharmacological-Data-to-Predict-the-Activity-of-Enyzme-Compounds-Using-Machine-Learning
This is my diploma thesis with the title "Analysis of Pharmacological Data for the Prediction of the Effectiveness of Cardio-Protective Activity" under the supervision of Dr. George Manis.
Impoved-LatexEditor
This project was made under the course : Software Development II regarding the Quality Check and Improvement of a "messy project".
Otsu-Adaptive-Thresholding
This code is about applying a adaptive threshold to a grayscale image using the Otsu method.
Covid-19-Analysis
Extended data analysis for Covid 19 pandemic. Data was provided by CDC (https://covid.cdc.gov/covid-data-tracker/#cases_casesper100klast7days)
Discrete-Cosine-Transform
This code is about taking the DCT of an image and applying 3 separate thresholds. The output is 3 new images with noise plus the calculation of each photo's PSNR.
Feature-Extraction-from-Breast-Cancer-data
In this brief project, i conduct an extented analysis on breast cancer data (data provided by https://www.kaggle.com/uciml/breast-cancer-wisconsin-data). I analyze the correlation between the features to extract the most valuable. I train 10 models with the initial features and then train the same models with the extracted features to show the effectivesness of the feature extraction methods used.
Geometric-Transformation
This code is about selecting a specific area in an image and the code will return this area to the user.
Google-Trends-Analysis
This small project is about analyzing the search data, based on Google Trends, during the Covid-19 pandemic for the year "2020". The study was conducted using woldwide searches and searches that were made in Greece.
Image-Quantization
This code is about applying reverse quantization to an image so we can add noise using the DCT and reverse DCT. The output is 3 images with different noises plus the PSNR and Entropy of each separate image.
Image-Thresholding
This code is about applying a threshold to a grayscale image.
Indexing-Rtree
This code is about creating an Rtree indexation system for a plethora of data. The data is in the form of spatial object( Mbrs ) and the algorithm used to create the Rtree is the "STR".
Kmeans-LVQ-Implementation
Simple Kmeans-LVQ (Learning Vector Quantization) implementation to a series of 2D vectors.
LatexEditor
The program creates a Latex Editor and provides a usefull UI.
Movies-Explorer
This project is about getting information about Movies, through a UI, that are loaded from the IMDB data bases.
Multi-Layer-Perceptron-MLP
Implemetation of Multilayer Perceptron (MLP) with 2 hidden layers and one output layer to a series of 2d vectors. The code classifies a series of data using a neural network, that we train by hand.
Applied-Affine-Transformation
This code is about doing a affine transformation in a grayscale image and using the method of nearest neighbor interpolation.
BraXaPsa-III
This program creates a CandyCrash like game called BraXaPsa III using the OpenGL/GLUT libraries.
Minecraft-Simulation
This project is about creating a Minecraft clone game working with the Unity3D platform.
Robot-Movement
This project is a more theoretical approach to a robot movement simulation that uses Euler Methods.
Top-k-Algorithms
This code is about implementing the HRJN algorithm in 2 sets of data and calculating the Top K couples that have the biggest sum of "weights". Additionally i created a changed version of the HRJN algorithm that does the same thing but with a difference in run-time.
CNN-vs-ANN-for-ECG-Heartbeat-Categorization
In this project i trained 2 deep learning models for ECG Categorization of 5 types of heartbeat. I showed how a deep model could be trained for real world data and also showed the difference between a ANN vs a CNN
K-Means-from-Scratch
In this short project i demonstrate the main idea and fucntion of the K-means clustering algorithm from scratch, using basic python code. The algorithm is applied on a set of 2D vectors that is provided with the notebook.