Building a chatbot with Rasa

Introduction Before getting stared with the development lets first dwell into the requirements and why we drilled down to the mentioned technology. I wanted to build a chatbot which is able to learn the intent of the user, interact intelligently, perform actions if users asks so, provide efficient learning mechanism and most importantly doesn’t use any… Read more »

Understanding the forecasting algorithm: STLF Model

Predicting the future based on historical information Introduction Lets start with understanding what is forecasting all about? Well its the best prediction of the future values provided the insights learned from the historical data. And as simple as it may sound, every forecasting algorithms tries to do so, alas with different assumptions. The STLF algorithm… Read more »

Reinforcement Learning with Multi Arm Bandit (Part 2)

Let make the problem a little bit more….complex! Recap This is in continuation of the original post, I highly recommend to go through it first, where we understood the intuition of multi arm bandit and tried to apply e-greedy algorithms to solve a representative problem. But the world isn’t that simple. There are some factors,… Read more »

Reinforcement Learning with Multi Arm Bandit

Click here to share this article on LinkedIn » Lets talk about the classical reinforcement learning problem which paved the way for delayed reward learning with balance between exploration and exploitation. What is Multi-Armed Bandit Problem? The ‘bandit problem’ deals with learning about the best decision to make in a static or dynamic environment, without knowing the… Read more »

Intro to Bitcoin — the economics and technology behind it

Introduction Bitcoin has been in news for some time now, be it for the mysterious founder Satoshi Nakamoto (nobody know who he/she is) or its ripple in the socio-economical world due to the ever fluctuating price or its underlining open source peer to peer blockchain technology. From major institutions shunning it, to countries trying to… Read more »

How I beat the online game 1to50 with neural network

The 1to50 bot — in all its glory! Well it all started when some of my friends started playing this online game 1to50 and they challenged me to beat their score. So the rule is quite simple, start clicking the blocks with numbers, in incremental order, as fast as you can. Seems quite easy, right!? It did to… Read more »

Reinforcement Learning with Q tables

Reinforcement learning — Agent’s action and environments reply Click here to share this article on LinkedIn » What is reinforcement learning Reinforcement learning is an area of machine learning dealing with delayed reward. What does this means? Well, simple, let me explain this with an example. For this I am assuming you have heard (better if you know)… Read more »