Harshvardhan Gupta

Harshvardhan Gupta blogs at Hacker Noon and tweets at @harvey_slash

5 Ways to Get Started with Reinforcement Learning

By Harshvardhan Gupta, Hacker Noon on September 20, 2017 in Deep Learning, Machine Learning, Neural Networks, Reinforcement Learning
We give an accessible overview of reinforcement learning, including Deep Q Learning, and provide useful links for implementing RL.
Using AI to Super Compress Images

By Harshvardhan Gupta, Hacker Noon on August 21, 2017 in AI, Compression, Convolutional Neural Networks, Image Recognition
Neural Network algorithms are showing promising results for different complex problems. Here we discuss how these algorithms are used in image compression.
DeepMind Relational Reasoning Networks Demystified

By Harshvardhan Gupta, Hacker Noon on August 15, 2017 in AI, Deep Learning, DeepMind, LSTM, Neural Networks
Every time DeepMind publishes a new paper, there is frenzied media coverage around it. We examine what is and is not real in recent work described as “DeepMind Neural Network Can Make Sense of Objects Around It”.
Train your Deep Learning Faster: FreezeOut

By Harshvardhan Gupta, Hacker Noon on August 3, 2017 in Deep Learning, Machine Learning, Model Performance, Modeling, Neural Networks
We explain another novel method for much faster training of Deep Learning models by freezing the intermediate layers, and show that it has little or no effect on accuracy.
Train your Deep Learning model faster and sharper: Snapshot Ensembling — M models for the cost of 1

By Harshvardhan Gupta, Hacker Noon on August 2, 2017 in Deep Learning, Ensemble Methods, Gradient Descent, Machine Learning, Model Performance
We explain a novel Snapshot Ensembling method for increasing accuracy of Deep Learning models while also reducing training time.
CAN (Creative Adversarial Network) - Explained

By Harshvardhan Gupta, Hacker Noon on July 12, 2017 in Art, Deep Learning, GANs, Generative Adversarial Network
GANs (Generative Adversarial Networks), a type of Deep Learning networks, have been very successful in creating non-procedural content. This work explores the possibility of machine generated creative content.

Harshvardhan Gupta

Latest Posts

Top Posts