With Great power comes great responsibility. — Pandas is a Data Analysis python library that aids in working with tabular data stored in spreadsheets and databases. It provides a vast set of functionalities for manipulating and transforming structural data aka dataframes. In this blog post, we shall discuss 3 simple tricks for speeding up Pandas operations. 1. Stop using iterrows() : Data…

Computing the steady-state behavior of a Markov chain using Python — Markov Chains Refresher: A Markov chain is a discrete-time discrete-valued random process that follows that Markov property. Mathematically, a Markov chain is denoted as Where for each time instant n, the process takes a value from a discrete set defined by

All about Markov chains and steady-state probabilities. — In addition to proving the central limit theorem, Andrey Andreyevich Markov played a pivotal role in developing Markov chains. Markov chains are used to model discrete-time, discrete space random processes with applications across multiple domains including Finance, Advertising, NLP, SEO, Physics, etc. …

Deciphering the impact of News, the “Intelligent Investor” way — If you like staying on top of your financial news feed, chances are you have stumbled on jazzed-up news articles that reveal “hot off the press” news related to some publicly-traded company. If you are an investor, chances are, you have, at some point, acted on the news by buying/shorting…

Discusses Implementation details with code — The goal of this blog is to dig right into the weeds of implementing a multilabel text classifier using Keras. If you have clicked on this story, I assume you are familiar with the problem formulation and how it differs from a traditional multiclass classifier. …

Gotchas to avoid while training a multilabel classifier. — In a traditional classification problem formulation, classes are mutually exclusive, i.e, each training example belongs only to one class. A binary classification problem involves classifying the dataset into two possible classes, eg: spam vs not spam, cat vs dog, churn prediction, etc. A multiclass classification problem involves classifying dataset into…

Writing computationally efficient python code using Dask: A distributed framework for Analytics — Why Dask ? Over the last couple of months, I have been working relentlessly to productionize a computationally expensive algorithm. The algorithm ingests millions of rows with unstructured text data and trains a language model before proceeding to the inference stage. One of the bottlenecks in deploying the feature was its inability to…

Pitfalls and considerations to be made while working with DS/ML features — Once upon a Time: In December 2019, Zillow started a house flipping hustle - “Offers”, allowing sellers to directly sell houses to Zillow thereby eliminating the lengthy bidding and closing process. Zillow saw this as an opportunity to monetize and profit from the rich real-estate data at its disposal. The “Offers” program was powered…

Diving Deeper into Deep Learning — ReLU vs Sigmoid Activation function. — Over the last few years, Deep Neural Network architectures have played a pivotal role in solving some of the most complex machine learning problems. Training a Deep neural network is not trivial and often involves billions of parameters trained over terabytes of data. Building a robust trainable network is an…