Pinned Repositories
Assembler
Basic-PCA
This is the second assignment in Machine Learning course, it it to get familiar with Principal Component Analysis and more familiar with python.
Binary-Classification-And-Evaluation
Coin-Detection-Circle-Detection-
In this Computer Vision assignment, I used filters to cartoonize any Image and familiarize with basic image processing, In the second part, Hough transform was used to identify coins as circles from images and plot them. Edges were detected first, and the using the detected edges and a pre-known radius, each edge point voted for the circles it might be contained in, with some thresholding and small optimizations for easier visual understanding, Each coin is recognized exactly.
Face-Recognition-Eigenfaces-
This is a basic face recognition program, It is trained on a small dataset and gave good results in identifying the people, PCA and LDA were used to reduce the dimensions and find the most common (Eigenfaces) and then a basic KNN classifier was used to predict the result of each test record. This project is to familiarize with PCA, LDA, KNN Classifier and Some Python libaries.
Familiarizing-with-concurrency
Handling races, synchronization and deadlock conditions. By using semaphores, I have solved the Producer-Consumer problem using a protected buffer due to the speed mismatch.
Huffman-Compression
C++ Code to compress any file. Reduces files by (average) 39% of their original size. (depends on how many different characters exist in the file)
Image-Panoramas
Forms a panorama/Image Mosaic given 2 images to the same place.
Spectral-Clustering
Spectral Clustering using k-way normalized cuts, The k-way normalized cut tries to cut the similarity graph into k partitions, It cares about making the new clusters of close weight, and, of course the cuts must be of minimum cost, so to cluster the data into k clusters, almost same size, and each cluster is closely related. In this project It is shown against good dataset to visualize the difference between this type of spectral clustering and a normal k-means clustering.
Topolgy-Mapping
Topology mapping is a partitioning technique that maps the simulated nodes to different physical nodes. In this assignment, we will use spectral clustering to partition a given network topology on the available physical nodes. The clustering technique should find the cut that minimizes the traffic between different partitions.
Mobad225's Repositories
Mobad225/Assembler
Mobad225/Coin-Detection-Circle-Detection-
In this Computer Vision assignment, I used filters to cartoonize any Image and familiarize with basic image processing, In the second part, Hough transform was used to identify coins as circles from images and plot them. Edges were detected first, and the using the detected edges and a pre-known radius, each edge point voted for the circles it might be contained in, with some thresholding and small optimizations for easier visual understanding, Each coin is recognized exactly.
Mobad225/Face-Recognition-Eigenfaces-
This is a basic face recognition program, It is trained on a small dataset and gave good results in identifying the people, PCA and LDA were used to reduce the dimensions and find the most common (Eigenfaces) and then a basic KNN classifier was used to predict the result of each test record. This project is to familiarize with PCA, LDA, KNN Classifier and Some Python libaries.
Mobad225/Image-Panoramas
Forms a panorama/Image Mosaic given 2 images to the same place.
Mobad225/Spectral-Clustering
Spectral Clustering using k-way normalized cuts, The k-way normalized cut tries to cut the similarity graph into k partitions, It cares about making the new clusters of close weight, and, of course the cuts must be of minimum cost, so to cluster the data into k clusters, almost same size, and each cluster is closely related. In this project It is shown against good dataset to visualize the difference between this type of spectral clustering and a normal k-means clustering.
Mobad225/Basic-PCA
This is the second assignment in Machine Learning course, it it to get familiar with Principal Component Analysis and more familiar with python.
Mobad225/Binary-Classification-And-Evaluation
Mobad225/Familiarizing-with-concurrency
Handling races, synchronization and deadlock conditions. By using semaphores, I have solved the Producer-Consumer problem using a protected buffer due to the speed mismatch.
Mobad225/Huffman-Compression
C++ Code to compress any file. Reduces files by (average) 39% of their original size. (depends on how many different characters exist in the file)
Mobad225/Topolgy-Mapping
Topology mapping is a partitioning technique that maps the simulated nodes to different physical nodes. In this assignment, we will use spectral clustering to partition a given network topology on the available physical nodes. The clustering technique should find the cut that minimizes the traffic between different partitions.
Mobad225/Clustering-Evaluation
Clustering Evaluation techniques including internal and external measures (F Measure - Jaccard index - Rand index - Purity - BetaCV)
Mobad225/Housing-Prices-Regression
Following the steps from the Hands-on Machine Learning 2 book, I understood and re-implemented this project, which is basically a thorough pass on what a usual Machine Learning Project is like.
Mobad225/Image-Steganography
Image Steganography sounded interesting, so I decided to hide the English dictionary in some grayscale image to check it out myself and see if I could spot differences between the photos.
Mobad225/LDA-and-K-means
Application to familiarize with Linear discriminant analysis and K-means. LDA reduces dimensionality of the data, but unlike PCA it tries to seperate data of different classes rather than just keep the dimensions which keep the most variance. The K-means algorithm is a basic clustering algorithm to cluster a set of data to K clusters, using the euclidean distance between the classes' discriminant (Class mean)
Mobad225/S-DCNet
Implementaion of S-DCNet (ICCV 2019)
Mobad225/Visualizing-Iris-Dataset
First Assignment in Pattern Recognition (Machine Learning) Course, It is basically a good visualization of the famous Iris dataset, I have also used the principle component analysis to reduce the dimensions of the data to easier visualize it (Keeping very good accuracy)