jayanath1987 Goto Github PK

ai-health

aie

aliensystems

attendance_npss

backupcronemail

Schedule SQL dump status email

backupcronemail-server

Email Backup Status Cron

build-a-cybersecurity-startup

This is the code for "Build a Cybersecurity Startup" by Siraj Raval on Youtube

computational-music

Generating Music Programmatically

dlnn-08

This week we discussed about Unsupervised Deep Learning Algorithms. Under the topic we started Autoencorders and implemented a model to denoise the noise in Handwritten Digits. The MNIST dataset with random noise was used.

dlnn-assignment-01

This Assignment is based on basic Python Programming Concepts

dlnn-assignment-02

Implementation of a simple FFNN for predicting the probability of having a heart disease

dlnn-assignment-03

dlnn-tutorial-02

dlnn-tutorial-03

In this tutorial we are going to discuss how to train the Tensorflow object detection Api to detect your custom object (example- Banana),

dlnn-tutorial-1

This tutorial explains how to build up a drawing canvas - web application using Python-Flask and JavaScript. The back end is implemented using Flask. At the end of the tutorial, you will be guided to the 1st in class project, Handwritten Digits Recognition app. A FFNN type neural network will be used for the project and the FFNN will be trained using the MNIST dataset.

dlnn-week-01

Introduction to Artificial Intelligence, Python Programming Basics and Essential Python Module

dlnn-week-02

In this week we discussed about Feed Forward Type Neural Networks, Supervised Deep Learning and Forward Propagation and implemented simple 4 layer Deep FFNN for Iris Flower example.

dlnn-week-03

This week we discussed about loss functions, loss optimization, gradient decent, Adaptive learning rate optmizers and back-propagation algorithm. Started the 1st in class project, handwritten digits recognition app using flask and keras. A FFNN type neural network was implemented and trained using the MNIST data-set.

dlnn-week-04

This week we discussed about Convolution Neural Networks. And started building up a simple CNN model for detecting cats and dogs.

dlnn-week-05

In this week trained a CNN to identify cat & dogs and tested it with some unseen data. We experienced that the CNN is suffering from over-fitting while training and ended up with a low validation accuracy like 75%. In coming weeks we will discuss about regularization methods for minimizing and avoiding over-fitting such dropout, early stopping, batch normalization and etc. As the 2nd In class project we implemented the NVIDIA self driving car model with Udacity Self Driving Car Simulator.

dlnn-week-06

This Week we discussed, how view and visualize trainable parameters (Weights and Biases) of a trained Neural Network. Then under applications of CNN we implemented the Tensor-flow object detection API. We applied used the mobile net coco version for this implementation.

dlnn-week-07

This week we discussed about Recurrent Neural Networks, Long Short Term Memory Cells, Back Propagation through time and how to apply LSTMs for time series/ sequence data. Then we implemented a simple random LSTM Neural Network to predict stock market values using a real data-set from AAPL.

dlnn-week-10

In this week we discussed about Reinforcement Learning. Under the topic, Q Learning and Policy based Learning sections were discussed. And we implemented a RL Agent for Cart-Pole environment available in Gym Module.

dlnn-week-11

This week we tried to implement a CNN type Neural Network for Energy Disaggregation. Energy disaggregation is the problem of separating an aggregate energy signal into the consumption of individual appliances in a household.

docker-php-5.3-apache

documentation

dump_restore

eden

Sahana Eden is an Open Source Humanitarian Platform which can be used to provide solutions for Disaster Management, Development, and Environmental Management sectors.. Please sign CLA when submitting pull requests: http://bit.ly/SSF-eCLA

ehealth-admission

Patient Admission

ehealth-api

API