habildamania Goto Github PK

array-1

array-2

binary-search-1

binary-search-2

cloudflare-2020-general-engineering-assignment

cloudflare-2020-systems-engineering-assignment

cloudflare_assignment

Image Data Store for Cloudflare Assignment

coda-assignment

Movie Review System Web App

coda_ids

IDS for Movie Review System

competitive-coding-1

competitive_coding_2

design-1

design-2

dp-1

dp-2

dp-3

gradient-descent-algorithm

Gradient Descent is an Optimization technique to Update the weight vectors in a classifier in order to reduce the error of classification. Theoretically, the gradient descent should terminate when there is convergence and the error becomes 0 on a set of weight vectors. But practically, achieving a goal like this is not possible for a large dataset. This code gives a basic idea of the working of a gradient descent algorithm on a toy dataset. It can be replaced with a dataset with of ones choice. The number of iterations are set to 5. Number of iterations can be toggled to achieve better results.

hashing-1

hashing-2

id3-decision-tree-algorithm

The original source code of this file belongs to Google Developers. The Source code provided by google developers is free to use and modify. I have modified the source code and added some ancillary features to it. The changed made by me are as follows

maiden

A Machine Learning approach for Intrusion Detection Using Ensemble

mysql-contact-list

A Web based Application to store your address book digitally. The number of entries in the address book are not limited. Perform operations such as Adding, Deleting, Modifying and Searching a contact by any combination of the contact field. Store all information about the contact. This web App is also Mobile Responsive and can be accessed from any mobile Device or Laptops, Desktops, etc. Use of Jquery, AngularJS, AJAX, CSS, HTML, PHP and MySQL to provide the best and most seamless Operation.

perceptron

This code shows the basic working of a perceptron. A perceptron is a single neuron which accepts 'n' inputs each associated with a weight vector 'wi' for every 'xi' input and a bias 'w0'. It performs one forward pass and calculates the sum of all wi*xi called the summing function and passes it to a sign function which predicts the class of the data instance. The sign function assigns class 1 to positive values and class 0 to negative values. After every forward pass, the error is calculated for the data instance and using stochastic gradient descent optimization technique, the weight vectors are updated so that error at output becomes 0.

precourse_1

question-answering-system

A Question Answering Bot using Deep Natural Language Processing Pipeline

s30-mock-interviews

sci-kit-learn-lab

This code shows how to create a neural network with multiple hidden layers using the sklearn library in python. The neural network is then trained on the car evaluation dataset and the classification report is displayed using confusion matrix and various measures such as precision, recall and f-score. Pre-processing of the data (such as categorical to numerical, standard scaling, normalization) is performed before training. Train-test-split is used to separate training and test data. Libraries such as Numpy, pandas, preprocessing are used for data processing.

software_testing_project

This repository is for the Software Testing Project implemented as part of our coursework for the CS 6367.001 class at University of Texas at Dallas

tutorials

xor-problem-using-neural-network

The XOR problem is very famous in the machine learning domain. An XOR truth table cannot be classified correctly by a single perceptron. This shows that XOR is not solvable by linear classifiers. However, XOR can be solved using a group of perceptrons also called as a neural network. This code will show you how this is done.