Keep it simple! How to understand Gradient Descent algorithm

Keep it simple! How to understand #GradientDescent algorithm #MachineLearning

  • The blue line gives the actual house prices from historical data (Yactual)

    The difference between Yactual and Ypred (given by the yellow dashed lines) is the prediction error (E)

    So, we need to find a line with optimal values of a,b (called weights) that best fits the historical data by reducing the prediction error and improving prediction accuracy.

  • So, our objective is to find optimal a, b that minimizes the error between actual and predicted values of house price:

    This is where Gradient Descent comes into the picture.

  • Gradient descent is an optimization algorithm that finds the optimal weights (a,b) that reduces  prediction error.
  • Lets now go step by step to understand the Gradient Descent algorithm:

    Initialize the weights(a & b) with random values and calculate Error (SSE)

    Calculate the gradient i.e. change in SSE when the weights (a & b) are changed by a very small value from their original randomly initialized value.

  • Adjust the weights with the gradients to reach the optimal values where SSE is minimized

    Use the new weights for prediction and to calculate the new SSE

    Repeat steps 2 and 3 till further adjustments to weights doesn’t significantly reduce the Error

    We will now go through each of the steps in detail (I worked out the steps in excel, which I have pasted below).


In Data Science, Gradient Descent is one of the important and difficult concepts. Here we explain this concept with an example, in a very simple way. Check this out.

Continue reading “Keep it simple! How to understand Gradient Descent algorithm”

Practical Deep Learning For Coders—18 hours of lessons for free

Practical #DeepLearning For Coders—18 hours of lessons for free

  • After this course, I cannot ignore the new developments in deep learning—I will devote one third of my machine learning course to the subject.
  • I’m a CEO, not a coder, so the idea that I’d be able to create a GPU deep learning server in the cloud meant learning a lot of new things—but with all the help on the wiki and from the instructors and community on the forum I did it!
  • Sometimes I feared whether I would be able to solve any deep learning problems, as all the research papers I read were very mathy beyond reach of simple intuitive terms.
  • But Jeremy and Rachel (Course Professors) believe in the theory of ‘Simple is Powerful’, by virtue of which anyone who takes this course will be able to confidently understand the simple techniques behind the ‘magic’ Deep Learning.
  • The course exceeded my expectations and showed me first hand how both Deep Learning and ourselves could change the world for better.

fast.ai’s practical deep learning MOOC for coders. Learn CNNs, RNNs, computer vision, NLP, recommendation systems, keras, theano, and much more! neural networks!

Continue reading “Practical Deep Learning For Coders—18 hours of lessons for free”

Practical Deep Learning For Coders—18 hours of lessons for free

Welcome to a 7 week course, Practical Deep Learning For Coders, Part 1,

  • The course exceeded my expectations and showed me first hand how both Deep Learning and ourselves could change the world for better.
  • If you can code, you can do deep learning
  • It is very hands-on and adopts a top-down approach, which means everyone irrespective of varying knowledge can get started with implementing Deep learning models immediately.
  • If you are looking to venture into the Deep learning field, look no further and take this course.
  • I now have the tools to apply deep learning models to real world problems.

fast.ai’s practical deep learning MOOC for coders. Learn CNNs, RNNs, computer vision, NLP, recommendation systems, keras, theano, and much more! neural networks!

Continue reading “Practical Deep Learning For Coders—18 hours of lessons for free”

Practical Deep Learning For Coders—18 hours of lessons for free

Practical Deep Learning For Coders

  • The course exceeded my expectations and showed me first hand how both Deep Learning and ourselves could change the world for better.
  • If you can code, you can do deep learning
  • It is very hands-on and adopts a top-down approach, which means everyone irrespective of varying knowledge can get started with implementing Deep learning models immediately.
  • I now have the tools to apply deep learning models to real world problems.
  • If you are looking to venture into the Deep learning field, look no further and take this course.

fast.ai’s practical deep learning MOOC for coders. Learn CNNs, RNNs, computer vision, NLP, recommendation systems, keras, theano, and much more! neural networks!

Continue reading “Practical Deep Learning For Coders—18 hours of lessons for free”

GitHub

.@mza announcing a new deeplearning benchmark:  #reInvent

  • To run comparisons in a deep learning cluster created with CloudFormation
  • The runscalabilitytest.sh script runs scalability tests and records the throughput as images/sec in CSV files under ‘csv_*’ directories.
  • Step 1: Create a deep learning cluster using CloudFormation .
  • Scalability Comparison Scripts for Deep Learning Frameworks

Contribute to deeplearning-benchmark development by creating an account on GitHub.
Continue reading “GitHub”

Google’s DeepMind AI Is Now Capable Of Self-Teaching New Things – The Merkle

.@Google DeepMind #AI Capable Of Self-Teaching New Things |   /@DeepLearn007 @themerklenews

  • Google, who developed DeepMind, recently bolstered their AI solution to make it learn new tricks faster.
  • Artificial intelligence capable of teaching itself new things can be seen as a troublesome development.
  • Google’s DeepMind AI Is Now Capable Of Self-Teaching New Things
  • Increasing the performance of this AI solution is of the utmost importance, even though its track record speaks for itself.
  • One of the primary selling points of artificial intelligence is how this technology can learn over time.

One of the primary selling points of artificial intelligence is how this technology can learn over time. Google, who developed DeepMind, recently bolstered their AI solution to make it learn new tricks faster. According to tests, DeepMind is now capable of learning close to 87% of expert human performance in games. This is an exciting development, although its real life use cases remain to be determined.
Continue reading “Google’s DeepMind AI Is Now Capable Of Self-Teaching New Things – The Merkle”