coelhonunes Goto Github PK

100dayspython

100 Days

backpropagation

In this exercise I will use backpropagation to train a multi-layer perceptron (with a single hidden layer). The experiment deals with different patterns and see how quickly or slowly the weights converge. Then it shows the impact and interplay of different parameters such as learning rate, number of iterations, and number of data points.

bmi

Udemy guided application

cnn-cifar-10-dataset

In this exercise I will build and train convolutional neural networks. In the first part, I walk through the different layers and how they are configured. In the second part, I built my own model, train it, and compare the performance.

coreml-idoggys

My own package that is able to identify which dog is in the picture.

cryptocurrencyprojects

This is tapping inside Coin Market Cap to display the latest cryptocurrency values.

deep-neural-network-techniques

From the wide variety of Deep Neural Network techniques, I have orchestrated a project that deals with the following topics: Convolution Neural Network (CNN), Transfer Learning, Data Augmentation, and finally, Recurrent Neural Networks (RNN). Specifically, I have built a CNN to classify images on a dataset, then illustrated the power of using the pre-trained models such as VGG16 and ResNet. In addition, techniques in using that dataset with data augmentation and then using the RNNs to classify the dataset are present.

deep_learning_spring_2024

faceproject

complete

fakedata

Here I have used the Faker import to create fake data and manipulated it in different ways.

flashchat

Udemy guided application

flixster

This app introduced Android Studio and linked data information through an Application Programming Interface (API)

handwritten-image-detection-mnist-data

In this exercise I will work with image data: specifically the famous MNIST data set. This data set contains 70,000 images of handwritten digits in grayscale (0=black, 255 = white). The images are 28 pixels by 28 pixels for a total of 784 pixels. This is quite small by image standards. Also, the images are well centered and isolated. This makes this problem solvable with standard fully connected neural nets without too much pre-work.

hotdog

Udemy guided application

imagefacedectection

This will count how many faces there are inside the imported picture.

javadatastructurepart1

This includes the following data structure techniques: exponentiation, recursive function, and sorted array.

javadatastructurepart2

This includes the following data structure techniques: binary search, binary search tree nodes, and one stack.

javadatastructurepart3

This includes the following data structure technique: Min Heap

javadatastructurepart4

This includes the following data structure technique: Radix Sort

linkedinalgorithms

livefacedectection

This application will box in each face that the camera can see and tell you how many faces it detects.

ml-az

Machine learning modules from Udemy

mustacheapp

A filter that places a mustache on the users face

myinstagram

This is an Android imitation of Instagram with live post, posting, and updates through APIs.

portfolio-html-css

This project involves raw HTML and CSS code to make a copy of my portfolio.

pre-trained-models

In this exercise I will show how to load pre-trained models such as VGG16 and ResNet. This is a fairly simple exercise designed to get familiar with models like VGG and Resnet and the output they give.

predictionregression

Variations in showing data through visual computation.

quantum-2024

readingfiles

This project reads text files and organizes the results into text files with serval parameters.

rnns-imdb-data

For this exercise, I will train a "vanilla" RNN to predict the sentiment on IMDB reviews. The data consists of 25000 training sequences and 25000 test sequences. The outcome is binary (positive/negative) and both outcomes are equally represented in both the training and the test set.