G. Kishan

G. Kishan

UnderGrad @ IIT Roorkee. Interested in interplay at physics and machine learning.

Sitemap

A list of all the posts and pages found on the site. For you robots out there is an XML version available for digesting as well.

Pages

Posts

Abstract of the paper “Competitive Gradient Descent”

less than 1 minute read

Published:

This blog explains method proposed in the paper Competitive gradient descent (Schäfer et al., 2019). This has been written as a supplimentary to the reproducibility report for reproducibility challenge of NeurIPS’19. The report has been published on arXiv. It can be seen at [https://arxiv.org/abs/2001.10820] with this github repository as its source.

Tricks for Manipulating Probability

less than 1 minute read

Published:

Dexterity in probability is needed to solve the fundamental problems of machine learning and artificial intelligence. This blog aims at summing up together the various techniques applied in different probability problems to make calculations easier and, sometimes,even possible!

portfolio

projects

Cat-Dog Classfier : Beginning with kaggle

Published:

  • This project is my biggining with kaggle. It contains an implemetation of basic CNN model in in TensorFlow.
  • Follow along the jupyter notebook on github to learn how to incorporate TensorBoard and debugging using it.

Blockchain

Published:

This project is created for having fun and learning some basics of Blockchain. It cointains a very simple blockchain implementation in Javascript. Follow the link to learn some basics and play around! It includes some features like:

  • Simple proof-of-work algorithm
  • Verify blockchain (to prevent tampering)
  • Generate wallet (private/public key)
  • Sign transactions

Self Balancing Bot

Published:

  • This project aims at stabilising a two-wheeled robot (naturally in non-equilibrium) using Reinforcement Learning.
  • A virtual bot is created using ROS URDF and PyBullet is used to simulate physics.
  • Q-learning is used as a Learning algorithm for the agent implemented using deep Q-Network, a deep neural network.
  • Click for details on hardware implementation and environment set up.

Raspberry Pi Object Detection

Published:

This repository contains python script for the object detection on Raspberry Pi in real time using OpenCV. It uses a already trained MobileNet Architecture stored as Caffe Model. This model uses Single Shot Detection(SSD) algorithm for prediction. It contains :

  • Guideline to run this project.
  • Abstract Code WalkThrough
  • Basic Learning Resources
    • Raspberry Pi
    • OpenCV
  • Fun Project for Contributors. Reference Materials
    • A blog from OpenCV
    • A video from sentdex (pythonprogramming.net)

PunyNet

Published:

  • This project is developed for MicroNet: Large-Scale Model Compression Competition, a model compression challenge hosted at NeurIPS’19.
  • We trained and quantized a resnet18 on CIFAR100 dataset.
  • The approach was of learning a codebook C that minimizes the difference between the output activations and their reconstructions.
  • Click for a detailed description.

Microsoft Workshop on Quantum Computing

Published:

  • This project consists of implementation (in Q# + c#) of several quantum algorithms discussed in the workshop.
  • This all work has been accomplished under the course “IPH-305: Introduction to Quantum Computing and Quantum Programming in Q#” in the supervision of instructor Prof. Ajay Wasan, Department of Physics, IIT Roorkee.
  • Click for a detailed description.

publications

teaching

Teaching experience 1

Undergraduate course, University 1, Department, 2014

This is a description of a teaching experience. You can use markdown like any other post.

Teaching experience 2

Workshop, University 1, Department, 2015

This is a description of a teaching experience. You can use markdown like any other post.