The Fifth Elephant 2018
2018-07-26..2018-07-27
Bangalore, India
[The Fifth Elephant 2013, The Fifth Elephant 2015, The Fifth Elephant 2016, The Fifth Elephant 2018, The Fifth Elephant 2019, The Fifth Elephant 2020, The Fifth Elephant 2023 Monsoon, The Fifth Elephant 2025 Winter Edition, The Fifth Elephant Pune edition 2026]
as the organisers put it
The Fifth Elephant is India’s best big data and machine learning conference. It is a conference for practitioners by practitioners.
who it's for
data scientists, data engineers, machine learning engineers, platform engineers, product managers, site reliability engineers
topics
linear regression pitfalls; classification model evaluation; food recommendation systems; data quality in large systems; structural estimation for mobility; scaling stateful data systems; ranking optimization for discovery; data-aware product design; streaming expressions for analytics; Spark scalability analysis; entity search across text and graph; serviceability under high demand
| title | author | length |
|---|---|---|
| So you think you know about linear regression ... | Chris Stucchio | 49m |
Explains how linear regression is a Bayesian generative model, where ridge regression = Gaussian prior and multicollinearity comes from a smeared likelihood… | ||
| A study in classification | Ramanan Balakrishnan | 30m |
Builds an ML pipeline that assigns Harmonized System (HS) classification codes using scraped international shipping records, third‑party aggregators and… | ||
| The power of intuition in data science, and why it will always have a role | Avi Patchava | 46m |
Argues that intuition is a learned system view that lets people imagine counterfactuals and spot causal links data alone misses. Gives concrete methods… | ||
| Our experiments with food recommendations | nitin hardeniya | 27m |
Swiggy describes running production experiments that move from restaurant‑first to an item‑first dish discovery (launched in Bangalore) using a hybrid of… | ||
| Compromising a $6B big data project through poor data quality: the Aadhaar case study | Anand V | 40m |
Claim: country‑wide offline enrollment creates a window for mass fraud. Enrollment clients are Java SE apps whose patched jars remove biometric and GPS checks… | ||
| Structural estimation methods to model user behaviour in bike-sharing systems | Ashish Kabra | 49m |
The cities of Paris, London, Chicago, and New York (among many others) have set up largescale bike-share systems to facilitate the use of bicycles for urban… | ||
| Why it is harder to scale with stateful, data-driven systems | Regunath B | 37m |
Claim: state never disappears it moves and scaling stateful services is harder. Flipkart Flux is an open source stateful serverless orchestrator that runs up… | ||
| Improving product discovery via relevance and ranking optimization | Akash Khandelwal | 55m |
In e-commerce, recommendations play a key role not only in customer satisfaction by improving discovery but also helps fulfill business objectives. In this… | ||
| Design for Data | Paul Meinshausen | 40m |
Argues that data generating process and workflow, not algorithms, determine success. Shows Housing.com reaching million properties but supply curves miss user… | ||
| Building analytics application with streaming expressions | Amrit Sarkar | 35m |
Apache Solr, an open source search engine project, has come a long way since its inception driving applications to have near-real time data mixed with… | ||
| Weaponizing data for politics | Shivam Shankar Singh | 30m |
We’ve all heard of the prevalence of Data Analytics in the political realm and stories of how companies like Cambridge Analytica influenced elections with the… | ||
| Sparklens: understanding the scalability limits of Spark applications | Rohit Karlupia | 39m |
Sparklens (OSS github.com/qubole/sparklens) uses Spark event listeners to compute a job’s critical path and ideal time and to simulate scaling effects; it… | ||
| Needle in a haystack : entity search on text and graph | Uma Sawant | 44m |
Web search today is moving towards displaying “answers” rather than making the user browse through pages to find what they want. “Entity” search queries, where… | ||
| Serviceability under high demand | Venkateshan K | 40m |
Defines a concrete metric stress ratio = undelivered orders / delivery agents and treats problem when s > 1; models the undelivered orders as a queue, uses… | ||
| nothing matches | ||