niloufaridani,Niloufar Faridani,github

adaline-madaline

In this project, two sets of two-dimensional data points are classified using an Adaline neural network. It shows how the network may separate initial datasets with certain mean and standard deviation values and and also compare it to the seperation that MadaLine network had done..

air-polution-prediction

This project is about predicting the PM2.5 with Pearson correlation and also with CNN+LSTM network

cnn-model-for-image-classification-on-mnist-fashion-ds

The exercise aims to use CNN architecture to classify the MNIST-Fashion dataset. Key tasks include dataset loading, designing two architectures based on a referenced article, explaining and comparing their layers, evaluating their performance, comparing SGD and Adam optimizers, and explaining the purpose of dropout layers in neural networks.

conditional-image-synthesis-with-auxiliary-classifier-gans

This project is implementing AC-GAN network with Normal loss and Wasserstein loss

decision-tree-algorithm

The project involves predicting Titanic passenger survival based on their attributes. Initially, a decision tree with a depth of 3 is used, and the impact of increasing depth is analyzed. The enhancements using Bagging and Random Forest techniques are proposed. Finally, a Random Forest with multiple decision trees is employed to improve accuracy.

deep-convolutional-gan

This project is implementing UNSUPERVISED REPRESENTATION LEARNING WITH DEEP CONVOLUTIONAL GENERATIVE ADVERSARIAL NETWORKS paper

fake-news-detection

This project is about detecting fake news using hybrid CNN+LSTM network

fourier-transform

This project is about implementing manual fourier transform function and Interpolation function.

frequency-estimation-and-fourier-of-periodic-signal

This project is about frequency estimation of a signal, using the furier transform and interpolation.

image-classification

This project is about image classifcation using MLP + BEiT with CIFAR10 dataset

knn-algorithm

In this question, we analyze the impact of different learning metrics on k-NN classification using the "wine" dataset. We explore two learning metrics assess the role of the 'k' parameter in these metrics versus standard k-NN, and consider data visualization in lower dimensions through PCA. We then evaluate data separation for various 'k' values.

mcclloch-pitts-neurons

The task involves building a binary multiplier using McCulloch-Pitts neurons. It takes two two-bit binary inputs and generates a four-bit binary output, utilizing a total of eight neurons. Each output's binary value will be determined, and a distinct network will be designed for each output based on its binary position.

multi-layer-perceptron

In this project, we explore housing price prediction using feature engineering and MLP models. We preprocess the data, visualize correlations, create and train MLP models with various configurations, and evaluate their performance by making price predictions on a test dataset.

multiple-classification-of-flower-images-using-transfer-learning

In this project, I implemented the Multiple Classification of Flower Images Using Transfer Learning paper

naive-bayes-algorithm

The objective of this task is to use a simple Naive Bayes algorithm to determine the accuracy percentage of digit recognition from a dataset where numbers 0 to 9 are manually written.

niloufaridani.github.io

Github Pages template for academic personal websites, forked from mmistakes/minimal-mistakes

problem-secratory

This problem involves finding the optimal strategy to maximize the probability of selecting the best candidate where decisions must be made immediately after each one. We compute probabilities using a formula with e/n for different k values to analyze these probabilities. We find the optimal rejection point that maximizes the prob of best option.

real-time-object-recognition

This project is about personalization of a new dataset about object recognition on YOLO-V6

resolution-impact-on-cnn-classification

The goal of this project is to compare the performance of CNN for classification on the 10-CIFAR dataset with different resolutions. The images have dimensions of 32x32 pixels.The task involves preparing three versions of the dataset with resolutions of 32x32, 16x16, and 8x8 pixels. Two different methods, TOTV and TVTV, is used for this purpose.

semantic-segmentation

This project is about implementing BEiT for semantic segmentation for scene_parse_150 dataset

transformer

This Hands-on project was about implementing the BERT model (Encoder, token embedding and position embedding)

niloufaridani Goto Github PK

Niloufar Faridani's Projects

Recommend Projects

Recommend Topics

Recommend Org